New Intel® Core™ Processors for Consumers

Formerly codenamed “Lynnfield,” these new chips are based on Intel’s award-winning Nehalem microarchitecture and are designed for consumers who need top-notch performance for digital media, productivity, gaming and other demanding applications. These processors, along with the new Intel P55 express chipset, are available today.All processors are lead- and halogen-free1 and feature Intel® exclusive Turbo Boost Technology. The top-of-the-line Core i7 processors also support Intel® Hyper-Threading Technology. Combined, these features give computer users absolute “intelligent” performance when necessary and optimum power-efficiency when the computer is lightly loaded.

Computers Just Got Smaller

The new chipset brings the most revolutionary design changes since the invention of the PCI bus in the early 1990s and sets the stage for Intel’s forthcoming 2010 compute platform. The Intel® P55 Express Chipset will be the baseline building block component for motherboards worldwide, delivering great new levels of performance and scalability for everyone from the retail buyer to the technically savvy do-it-yourselfer.

The new Core i7 and i5 processors are the first Intel processors to integrate both a 16-lane PCI Express 2 graphics port and two-channel memory controller, enabling all input/output and manageability functions to be handled by the single-chip Intel P55 Express Chipset. Previous Intel chipsets required two separate chips. A new Direct Media Interface (DMI) connects between the processor and chipset. The chipset supports 8 PCI Express 2.0 x1 Ports (2.5GT/s) for flexible device support. Dual graphics cards are supported in a “2×8″ configuration. The chipset also supports 6 SATA 3 Gb/s Ports with Intel® Matrix Storage Technology providing RAID levels 0/1/5/10. Up to 14 USB 2.0 Ports can be supported with the chipset’s integrated USB 2.0 Rate Matching Hub, along with Intel® High Definition Audio for premium digital sound. The new processors are the first to be supported by the new Land Grid Array (LGA) 1156 package and socket technology.

Better Entry Servers

Small businesses requiring 24/7 operation and educators now have more reasons than ever to buy a purpose-built server with Intel’s new Xeon processors and Intel® 3400 and 3420 chipsets. These new products improve small business productivity by running email, file, print and dynamic Web serving tasks more efficiently. They also improve education by enabling dependable classroom collaboration and making school administrative services more productive. Servers based on Xeon 3400 processors provide more dependability over desktop systems through differentiated features such as Error Correcting Code memory and RAID 0/1/5/10 for server operating systems. They are designed to help small businesses grow by enabling up to 64 percent2 more sale transactions and up to 56 percent faster business response time. This improvement is enabled with Intel’s Nehalem microarchitecture and a 4x improvement in memory capacity (32 GB). Intel® Turbo Boost Technology and Intel® Hyper-Threading Technology enable these servers to automatically adapt their performance to unique business needs. The processors launched today also include the Intel® Xeon® L3426, a low-power variant that delivers up to 188 percent improvement in energy efficiency per dollar than the previous-generation Intel® Xeon® X3380, and enables innovative server form factors for space and thermally constrained environments. For more information on these products, please visit www.intel.com/xeon.

Processor Number	Base Clock Speed (GHz)	Turbo Frequency (GHz)	Cores/Threads	Cache	1K Unit Price	Intel Hyper-Threading Technology?	TDP
Intel® Core™ i7-870	2.93	Up to 3.6 GHz	4/8	8 MB	$562	Yes	95W
Intel® Core™ i7-860	2.80	Up to 3.46 GHz	4/8	8 MB	$284	Yes	95W
Intel® Core™ i5-750	2.66	Up to 3.2 GHz	4/4	8 MB	$196	No	95W
Intel® Xeon® X3470	2.93	Up to 3.6 GHz	4/8	8 MB	$589	Yes	95W
Intel® Xeon® X3460	2.80	Up to 3.46 GHz	4/8	8 MB	$316	Yes	95W
Intel® Xeon® X3450	2.66	Up to 3.2 GHz	4/8	8 MB	$241	Yes	95W
Intel® Xeon® X3440	2.53	Up to 2.93 GHz	4/8	8 MB	$215	Yes	95W
Intel® Xeon® X3430	2.40	Up to 2.8 GHz	4/4	8 MB	$189	No	95W
Intel® Xeon® L3426	1.86	Up to 3.2 GHz	4/8	8 MB	$284	Yes	45W

− $4.2 billion state-of-the-art semiconductor manufacturing facility expected to create up to 6,400 new jobs and provide foundation for building AMD’s future technology innovations –

− Sen. Schumer, U.S. Rep. Murphy and Gov. Paterson participate in event −

Malta, N.Y. – July 24, 2009 – AMD (NYSE: AMD) today joined GLOBALFOUNDRIES and federal, state and local government leaders to mark and celebrate the official start of construction of Fab 2, a state-of-the-art semiconductor manufacturing facility at the Luther Forest Technology Campus (LFTC) in Upstate New York. Representing the largest public-private sector industrial investment in New York State history, the project is expected to be the most advanced semiconductor manufacturing facility in the world when it begins production, currently scheduled for 2012.

“Fab 2 represents the culmination of AMD’s multi-year vision to bring a leading-edge semiconductor manufacturing facility to the United States. This is an important opportunity to create thousands of jobs and strengthen U.S. competiveness in the high-tech industry,” said Dirk Meyer, president and CEO, AMD. “GLOBALFOUNDRIES is expected to be an ideal technology partner for the entire semiconductor design industry, providing cost-effective access to manufacturing capacity on an unprecedented scale. The multi-billion dollar investments in research and development and capacity expansion that GLOBALFOUNDRIES is planning, strengthen its position as a premier leading-edge semiconductor manufacturing foundry and make it the ideal AMD technology partner to help in bringing our innovative products to market.”

GLOBALFOUNDRIES was created as a joint venture by AMD and the Advanced Technology Investment Company (ATIC) on March 2, 2009 and marks the culmination of AMD’s Asset Smart strategy.

Supporting Resources

· Factsheet on AMD and GLOBALFOUNDRIES relationship
· GLOBALFOUNDRIES website with additional details on Fab 2
· Official timeline for Fab 2 construction

About AMD

Advanced Micro Devices (NYSE: AMD) is an innovative technology company dedicated to collaborating with customers and technology partners to ignite the next generation of computing and graphics solutions at work, home and play. For more information, visit http://www.amd.com.

Intel has been described as a 10,000 pound charging rhino and with good reason, they are a large company with deep pockets and access to some major resources. So when they move you can bet they build up some terrific momentum. During the years of the Pentium 4 they were caught grazing by the small and nimble AMD who snuck up on them with the Athlon and Athlon 64 CPUs. But once alerted, very much like that 10,000 pound rhino, they began to move, it was slow going at first and very hard to change directions once moving, but in the end they got all of that momentum and force aimed at getting the performance title back. This was handily done with the Core 2, but there was no slowing Intel down once moving. The charged right past a startled AMD who might not have even caught a glimpse at who blew past them, but they felt the wake for sure. Next up was Intel being first to the plate with the Kentsfield Quad Core CPU; this, while not a “true” quad core still had power and performance to spare. Now Intel has altered the rhino’s direction again. We finally see what the white coat wearing labs guys at Intel have been cooking up. So we bring you our 64-Bit evaluation of Nehalem, now called the i7. We will not go too deeply into the technical explanations of what Nehalem brings to the table but will concentrate on performance evaluation along with heat and power consumption.

Product: Intel Core i7  
Author: Sean Kalinich
Reviewed on: October 15th 2008
Product cost: See Below
Manufacturer: Intel
Spelling and Grammatical editor: Planetx64 Staff

{mospagebreak title=A CPU Divided}
What’s New?
The i7 is a new piece of silicon, it measures 263mm^2 and has over 731 million transistors. The i7 is larger than the 45nm QX9770 (214mm^2) but smaller than the QX9650 (286mm^2). What is interesting is that the i7 has less transistors than the QX9770 about 89 million less (QX9770 820 million transistors, i7 731 million transistors) but more than the QX9650 (731 Vs 582). The i7 is based off of Penryn (the P6) and even has the same pipeline size. It does have improvements in efficiency and power management yet Intel says that the Nehalem is the biggest architectural change to date.   So where are the principal differences?
The Dividing of the CPU:
Intel  has made a new distinction in the way a CPU is laid out. There are now two parts to the CPU. The Core and the Un-Core; the Core, as its name implies, consists of the execution units, L1 and L2 Cache, a memory ordering unit, paging, branch prediction and instruction reordering, scheduling and retirement functions. In a Quad Core i7 you will have four of these on the processor die.
The Un-Core is just about everything else; in the un-core you have the memory controller, L3 Cache, CPU I/O controllers and the QPI (QuickPath Interconnects)
Now this is not just a logical division but is also there for power management as the core and Un-core run in their own separate power planes.
{mospagebreak title=Power Management}
Power Management:
Since we are talking about power management; i7 brings something new to the Intel line up, this is in the form of:
An integrated Power Control Unit (PCU) and sensors
The replacement of clock gates with power gates
Reworking the L1 and L2 cache to increase the number of transistors from 6 to 8
Switching from Dynamic Domino Logic to static CMOS design. 

Now to understand all of this change I have to delve into each one a little bit.
The PCU that Intel has integrated into the Nehalem is unlike anything they have used before. This piece of CPU real-estate is about the same size as a 486 CPU (transistor count wise) has its own firmware and does nothing but monitor the CPUs power consumption, system requests, temperature, and CPU current and can makes its own decisions about power for the CPU despite what the OS may say, this prevents the rapid on/off power requests that can happen with OS driven power management. This PCU works in combination with the Power Gates, these new devices are exclusive to Intel and can actually shut off power to an inactive core while keeping the Core portion of the CPU on the same power plane. So like AMD’s Phenom you have each core able to operate at independent clocks and power usage, but unlike Phenom the Nehalem can almost completely shut off power to an inactive CPU Core. To further increase the difference, Nehalem can redirect some of that voltage to individual cores (in limited amounts) to improve single threaded performance. So this means that if only one core is active and getting close to being overworked the i7 can bump up the power to that core to give it a little boost. Intel calls this Turbo Mode. And can be adjusted by the number of cores active.
The move from 6T to 8T SRAM cells in the L1 and L2 cache was to reduce the amount of voltage needed to run these levels of cache. The next power saving feature was the shift from dynamic domino logic to static CMOS. Static CMOS uses less power than Domino logic and although there is a small performance hit from using it Intel has managed to almost negate this by improvements in some of the CPU algorithms and is still an overall improvement for the CPU that helps reduce the overall amount of power needed for the CPU to run.
{mospagebreak title=Memory and Cache}
Cache:
The Cache structure on the i7 has also changed, gone is the large L2 cache that was present with the Core 2 Now we see a return to smaller L2 cache (256k per Core) and the inclusion of a large (8MB) L3 fully-shared inclusive cache. And there have also been a few changes in each level of cache. L1 is slightly slower at 4 cycles Vs 3, L2 is no longer shared and is part of each core. The L3 cache is similar to AMD’s L3 cache in that is it shared between all of the cores of the CPU. By making it inclusive Intel has reduced the amount of core snoop traffic as the CPU needs to only look in the L3 cache to see if an instruction is present in cache (L1 and L2 are replicated in L3).
Internal Memory Controller:
One of the biggest differences is the inclusion of a memory controller into the CPU die, now AMD has been doing this for years while Intel has said it was not needed, but as the FSBs increased we saw memory performance decrease. This was the impetus that lead Intel to make this change; as it stands now even with a QX9770 and 2000MHz DDR3 you are not seeing the full potential of your CPU and RAM due to limitations of the older architecture. Intel’s new memory controller also breaks the conventional standard of Dual Channel memory; Nehalem’s is a Tri-Channel memory controller and can allow for up to three DIMMs per Channel this helps to improve memory performance and to get the most out of higher speed DDR3 memory. Intel estimates about a 40% improvement over current Dual Channel DDR3 FSB systems.
QPI (QuickPath Interconnect)
Intel’s QPI is very similar to AMD’s HyperTransport.  The QPI is a bi-directional connection or pathway to replace the old and outdated FSB.  It is capable of up to 6.4GT/s. Each link is 20bits wide with 2 bytes (16 bits) of effective bandwidth this gives it roughly 12.8 GB/s in each direction or 25.6GB/s per link total.
And I have said it before, Intel now gets to reap the benefits of AMD’s early adoption of this type of interconnect. They will not face the same performance and stability with applications from adding the memory controller to the die and using an internal interconnect like AMD did.
{mospagebreak title=HyperThreading}
The Return of HyperThreading:
In the old Networst (sic), P4 days HyperThreading was something of a joke. It was inefficient, power hungry and in most cases a complete waste. So why bring it back? To understand why it failed before we have to take a very quick look at what it does. HyperThreading was the friendly name for SMT or Symmetric Multi-Threading. This allows each physical core to appear as two logical core and therefore request additional instruction threads from the system (two per CPU core). With Networst and the P4 the pipelines were too long, the paths were not wide enough, CPU cache too small, and the memory bandwidth was nowhere near enough to make this viable.
Now with Nehalem we have dramatically increased memory bandwidth, larger Cache, short wide pipelines, new and deeper buffers to handle the extra processing load. With the i7 we see a CPU that is capable of efficiently handling the extra load that HyperThreading brings to the table.
Another big change is the core bus speed the i7 runs at. While with Core 2 we saw an eventual 400MHz the i7 has a much lower core bus of 133Mhz CPU and memory speed are adjusted by multipliers and by the speed of the QPI for each chip.

{mospagebreak title=One Last Thing}
One Last Thing on Change:
The Last change is not just to the CPU but to the platform. The i7 has changed sockets to handle the IMC (integrated memory controller)as well as other CPU to mainboard I/O paths and jumped from 775 pins (or lands) to 1366 pins. This means new Chipset and mainboard time. For testing Intel sent out an X58 based desktop board in the form of the DX58SO or Smackover (I think SmackDown is copyrighted so they could not use that). The feature of the DX58SO will be covered in detail in a later evaluation. But it is a very different design and implementation that what we have become used to from Intel. Both Asus and GIGABYTE have x58 boards ready and we will be talking about those in depth later too.

{mospagebreak title=Performance – The Rigs}

Performance:
As I mentioned above I did not want to spend too much time discussing the architecture and new features, if you are really interested go and take a look at the white papers and IDF documentation on Intel’s site. What I want to show is the performance of the i7, and in particular what you can expect from it in a 64-bit OS (Vista x64 Ultimate). I also wanted to be able to show if i7 represents a viable upgrade.
When AMD released the AM2 there was no performance gain to be had over the 939 Athlons. This presented a problem for many AMD users. So by the end of my testing I want to be able to tell you if i7 is a smart choice of if you should just stick with the Core 2 line for now.
Test systems
Intel  i7:
Core Extreme i7 -965 (3.2GHz 6.4GT/s QPI)
Core Extreme i7 -940 (2.93Ghz 4.8GT/s QPI)*
Core i7 -920 (2.66GHz 4.8GT/s QPI)
All core models are 130W TDP and have 8MB shared L3 Cache.
Intel DX58SO desktop mainboard (BIOS SO2260)
QIMONDA 3 x 1GB DDR3 1067 CL
Foxconn 9800GTX-512N
Intel X25 80GB SSD
Sony SATA DVD-RW
Mushkin 1000Watt PSU
* 940 performance was simulated by down clocking the 965 to 940 specifications (4.8GTs QPI and adjusting the multipliers)

Intel Core 2 Quad
QX9770 (3.2GHz, 1600MHz FSB, 12 MB L2 Cache)
Asus P5E64 WS Evolution
Kingston KHX13000D3LLK2/2GX (2x1GB DDR3 1625MHz @ 1600MHx 7-7-7-20)
Foxconn 9800GTX-512N
Intel X25 80GB SSD
Sony SATA DVD-RW
Mushkin 1000Watt PSU
Corsair Nautilus 500 used for cooling

Intel Xeon (the V8)
2 x Intel Xeon X5365 Processors (Quad Core 3.0GHz 1333MHz FSB)
Asus Z7S WS
4 GB (2x2GB) Kingston 667MHz FBDIMMs
Intel X25 80GB SSD
Foxconn 9800GTX-512N
Sony SATA DVD-RW
Cooler Master UCP 1100 PSU

AMD
AMD Phenom x4 9850 (Clocked to 9950 2.6 GHz)
Asus M3A32-MVP (BIOS 1203)
2GB (2x1GB) Kingston KHX9200D2K2/2G (@1066 5-5-5-18)
Foxconn 9800GTX-512N
Intel X25 80GB SSD
Sony SATA DVD-RW
Mushkin 1000Watt PSU
Corsair Nautilus 500 used for cooling

All systems were running Microsoft Windows Vista x64 Ultimate
nVidia driver version 178.24 was used for all testing.
I want to make a note here about my choice of GPU, at the time of writing it was the most powerful nVidia GPU I had on hand. I did have a pair of ATi HD4850s but there was a flaw in the available drivers that caused the DX and OpenGL subsystems to lag ad show about ½ expected performance unless the tests were run back to back 5-6 times. I contacted Intel who forwarded the information to ATi. As of publication I have not heard from ATi about a fix for this.  I hope they address this issue quickly or it will hurt their sales with this new platform.
* at the time of publication Catalyst 8-10 fixes the performance issue, however there was not enough time to re run all tests with the ATi HD4850.

{mospagebreak title=Performance – Applications}

Performance – Applications:
Let’s talk about application performance, the one thing that everyone probably wants to know is if the i7 is going to give new levels of performance over the existing Core CPUs or if (like the AM2) it is just a bump and one that can be easily bridged with a little over clocking.  Well with our usual tests in hand we have a combination of synthetic and real world tests to show you that the i7 is not just a bump in performance.

FutureMark:
From FutureMark we bring PCMark Vantage into our application testing pool, although a synthetic benchmark it still gives us a decent baseline for comparison and makes the numbers crowd happy.

Wow, that is really all I can think to say about the way the i7 simply outperforms everyone in 64-Bit performance. For 32 Bit performance only the 920 lags behind the QX9770.

Sisoft Sandra 2009:
Another numbers synthetic, Sandra 2009 is good for giving RAW performance numbers and also for giving a baseline for comparison.

The numbers here are interesting, it seems for sheer CPU power on the V8 can out do the i7. For Multi-core efficiency and memory bandwidth the i7 cannot be touched.

Everest 4.6:
The third in our synthetic test group I only used Everest for backing up the memory performance numbers for Sandra 2009, although it is useful for much more than that.

Again we see the tri-channel memory bringing the speed back to DDR3.

Cinebench R10 x64:
The last of our truly synthetic tests is Cinebench R10 x64, this test (From Maxon the makers of Cinema 4D) is intended to give a fair evaluation of your systems ability to render images and to support real-time OpenGL rendering.

Again the i7 stands head and shoulders above all but the V8 monster.

Photoshop CS3:
Photoshop is the industry standard for image manipulation. For my testing I used Driverheaven.net’s Photoshop CS2 benchmark.

Again we see the i7 965 at the top of the heap while the QX9770 come in second place, followed by the i7 940 and 920.

HyperPi 0.99b:
Although not truly a real world test HyperPi is still not a completely synthetic test. It is also perfect for testing memory to CPU to HDD performance and stability. I ran a 32M run on all cores for each system.

Here the numbers are interesting, the i7 965 (18:30:172 best core time) is at the top of the heap but coming in at a close 2^nd is the QX9770 (19:00:204 Best Core time) beating out the i7 940 by almost 50 seconds (19:50:102 Best core time).

Lightwave 9.3 x64:
Lightwave is one of the industry standards for 3D animation and a great test of system performance. Lightwave is also very CPU dependant, but also needs a fast HDD and memory for best results.
For testing I used the Moonbase Scene and rendered frame 32 @ 1920×1080 (HD 1080i) with 7-Pass PLD for AA, Gaussian Sharp reconstruction filter and classic cameras. Segment memory was set to 512MB.

And the King is dethroned, previously the V8 system was the fastest Ligthwave rendering system I have ever worked with (yes there are faster out there but they are massive systems) However we have a new speed king here. The i7 965 was able to outperform the V8 by 9 seconds. This might not sound like much but over a 60 frame render it is 9 minutes saved on your render time.

AutoGK 2.45:
For my Transcoding/encoding tests I used AutoGK 2.45 to transcode a standard 2-1/2 hour movie from DVD to 100% quality AVI. AutoGK2.45 uses virtual dub for its transcode but is more efficient and can spread the work load over multiple CPUs (cores) making it a good test for multi-core performance.

For Transcoding all of the i7 CPUs simply walked away from the others in the group. This is due mainly to the faster memory speeds and also better SSE code in the i7.

Application Performance Summary:
We had posed the question; is the i7 going to be a leap or a step in terms of performance. Well it seems the answer is, for the most part, a resounding yes for application performance. The i7 brings a better system to the table with its tri channel memory support IMC (internal memory controller) better caching structure and improved hyperthreading (over the P4). While the Massive V8 system still holds the crown for many of the synthetic CPU tests, there were only a few places where the entire i7 family did not win out over the competition. Even the lowest clocked i7 was able to outperform the rest of the crowd in Transcoding, and (other than the V8) in Lightwave as well. Looks like the i7 is will handle just about all of the apps you throw at it.

{mospagebreak title=Performance – Gaming}

There have been rumors and speculation around the Internet that the new i7 is not meant for gaming.
That Intel developed it for use in workstations and more professional systems. Well we are going to take a look and see if there is any truth to those rumors or if the i7 really can bring the same performance jump as we saw with our application testing. To find out we ran five current games at the highest playable resolutions (or max resolution for our 24” monitor) and also at minimal settings (1280×1024)
each game was played through 3 times and the median numbers recorded for minimum, maximum and average frames per second. Settings for each game at high hi-res and low-res are shown in each game section. All frame rates were recorded with FRAPS 2.9.

FutureMark – 3DMark Vantage:
3DMark Vantage like PCMark is not a real measure of performance, it is simply too easy to fool this test. A perfect example of this is the integration of the Agiea PhysX components into the nVidia GPU driver package. This gives an artificial bump to the CPU score which is not present when running an ATi card. Now I would not have a problem with this if it were to bump up the GPU score, but it is a false representation of what your CPU is actually capable of.

As you can see here 3DMark performance is affected very little by the CPU, and this is another reason that synthetic test can be misleading. In real gaming AI, physics, and depending on your choice of Audio Card/Codec, sound all play into gaming performance. Sound is especially important as sound and video are usually coded to stay in sync (or the game would be unusually odd to play). Even so we can still see that the i7 show it has power to spare for gaming, bear in mind that the CPU scores are artificially inflated by the nVidia PhysX package in the drivers I used.

Assassin’s Creed –DX10:
Assassin’s Creed can be a rough game on a system, it does have some rather bland AI but over all its graphics and audio tend to slow your game play down if you do not have a powerful enough system.
For my testing I ran the opening level from the fight in the courtyard through to your “death”.
Settings are shown below.

Hi-Res:
Under most high resolution gaming the CPU takes a back seat to the GPU. I found it interesting that in Assassin’s Creed the CPU made a rather big difference. The i7 965 showed a 20 frame per second lead over the older generation of CPUs. The 940 also showed a healthy performance lead while the 920 performed as well as the QX9770.

Low-Res:
Low resolution gaming can help take the GPU out of the equation when looking at raw CPU performance. For Assassin’s Creed we see the i7 take another healthy lead.

Crysis –Warhead:
I am not a big fan of Crysis or Warhead, but as Warhead is a popular and current game I wanted to include it. According to Crytek it is optimized for multi-core CPUs and Multi-GPU play. This is great but Multi-Core can have different levels of compatibility, and what might play well on a dual core CPU might play horribly on a quad. For testing I used the “Call me Ishmael” level from after the EMP blast through to the end. Settings are shown below:

Hi-Res:
In Crysis the i7 runs a little faster than all of the other CPUs tested granted not a big one (less than 4 FPS), even the entry level i7 920 shows a 1 FPS lead over the QX9770.
Low-Res
The i7’s lead is more evident in our low resolution testing with the 965 having a 19 FPS lead over the QX9770 and even the 920 was able to out the QX9770 by just over 3 FPS.

Lost Planet Extreme Conditions –DX10:
Lost Planet is an older game (released in January 2007) and was one of the first DX 10 and quad-core optimized games. It still can put a hurting on your system due to the sheer amount of fog and particle and other effects in play. For testing I ran through the complete first mission. Settings are show below:

Hi-Res:
Lost Planet was interesting, for some reason the AI seemed sluggish, game play was smooth at the resolution and setting chosen but it was not the same game. This was reflected in a lower frame rate than the Core, Phenom, and the V8, the fact that the V8 was able to run faster than the i7 leads me to think that Lost Planet does not like the i7’s hyper threading.  Since it was designed to run on a quad core CPU it might not like the extra thread per core or indeed be able to utilize them without losing performance.

Low-Res:
When we remove the GPU from the equation here we do see the 965 jump out in front while the 940 and 920 get good numbers. However these numbers were not enough to out run the Phenom 9850 (@ 2.6GHz) or the V8. In this test the QX9770 lagged behind the rest of the pack.

Call of Duty Modern Warfare:
Call of Duty 4 is a great game with good visuals and a large and massive “bar fight” AI system. While Call of Duty 4 is not going to slam your GPU that massive AI can bog down your CPU and make the game feel sluggish. For testing I ran the “Crew Expendable” level from your drop off on the ship to the end.

Hi-Res:
Performance here is very similar across the board with only the QX9770 lagging behind by 12 FPS.

Low-Res:
Here the differences are more apparent but still there is less than a 16 FPS difference from first to last place.

Bioshock:
Bioshock is a creation of the same minds that brought us System Shock 2 and features good (although a little cartoonish) graphics. AI is a little light as most of your opponents follow easily recognizable patterns. It is still a great game to play and good for a basic DX10 test. For my testing I ran the “Welcome to Rapture” level. Settings are shown below.

Hi-Res:
In Bioshock the numbers were all over the place with the 965 coming out on top and followed by the V8 and the 940. The QX9770 and i7 920 tied for fourth place while the Phenom lagged behind.

Low-Res:
In our low resolution testing we still see the 965 out in front but this time the V8 is in second place followed by the Phenom.

If a game can only run on one core, then you are back to raw horse power. Intel has not overlooked that possibility though as the i7 has plenty of horse power to spare. As you can see from our limited testing above the i7 even at high resolutions is able to give an edge in many games due to it new design. We see almost a reversal of the Pentium 4 Vs the Athlon 64 days when AMD was able to do the same thing to Intel in most games.

{mospagebreak title=Power and Heat}
Power and Heat:
Power and Heat are the bane of most computer systems. If you are an enthusiast you might tell yourself you are willing to accept the extra power usage to gain those extra FPS, 3DMarks, etc. However, when the power bill arrives you might sing a different tune. Heat is the slow death of a system, the more heat your system produces the shorter its life often is. The i7 is supposed to have a decent TDP (130Watts for the i7-965) but does that equal a power savings when placed in a system? To find out we measured each of the test systems power draw from the wall under the following:
Idle
Gaming (Lost Planet DX10)
Lightwave rendering
Transcoding
HyperPi (100% CPU for extended periods)

Lowest Draw/Peak Draw

As you can see only the V8 system used more power under load than the top-end i7 965. The 920 did better but was still not under the draw of the QX9770 or the 9850. Only at idle did the i7 CPUs show a lower power draw.

The heat output of the i7 is a little more than I would have expected. I am not sure why but I had expected it to have less heat production that it did. The ThermalRight 120 Xtreme did a great job in keeping the 965 cool, but the stock HSF did not show what I would call stellar performance.

{mospagebreak title=Value}
Value:
The Core i7 CPUs will fall into the same price lines as the existing Core 2 Quad CPUs. The i7 965 will hold the top slot at $999.00 the 940 will run $562.00 and the i7 920 will set you back $284. This is not too bad of a deal considering. However where you will take a hit in the wallet will be in purchasing an X58 board. The first generation of these will start out pricey to say the least and I would not expect them to drop anytime soon. Does this mean you should wait to get an i7? Not really, when looking at this upgrade it is important to remember that while you will spend about $100 more on a new “top-of-the-line” X58 board you will save a little on memory (DDR3 1066 is cheaper than DDR3 1600+)while the CPU should run about the same as what you would pay for a Core 2 Quad.  Existing HDDs GPU, PSUs and other components will still work (with the obvious exception of the HSF) I would not recommend using any current high speed DDR3, the risk of damage to the CPU is too much as many of these kits default to 1.8-1.9v and most boards will be set to auto right out of the box.  It is also important to remember when looking at the value of the i7 that even the entry level i7 920 out performs the top dog QX9770 in many cases. To wrap it up, the i7 will be an investment but one that is worth it based on performance per dollar.

{mospagebreak title=Conclusion}
Conclusion:
There is not much more to add to the i7, the performance really says it all. Intel has produced a true CPU evolution here, taking the Core marchitecture one step further and giving us a jump in performance. The fact that a single quad core CPU can in many cases perform as well or just behind a true Dual Socket 8 physical core system just by adding in one extra thread per core is impressive. The move to an internal memory controller designed for tri-channel DDR3 lends even more performance to the i7 system. There really was not a power savings to be seen over the older Core 2 but this may come with a die shrink to 32nm next year. The rumors and speculation that the i7 is not a gaming CPU have been dispelled in my mind. Even though gaming is highly GPU dependant, in only one game (Lost Planet) did I see the i7 truly lag behind the older Core 2 Quad. My guess is that Lost Planet, having been designed for a traditional Quad Core, did not like dealing with the extra thread per core and caused a performance loss. Application performance was nothing short of amazing as the i7 was able to render faster than a system with eight physical cores and lead the pack in everything but synthetic tests where the V8 remained out in front.
it would appear as if Intel has again launched another Conroe into the market and as I said when Conroe hit, AMD had better have something pretty amazing up their sleeve and be able to get it to market on time or they are in more trouble.  

Discussion Link

Introduction:
AMD, once the undisputed speed king, has in the past year suffered one setback after another.
First there was the lackluster release of the AM2 dubbed an evolution not a revolution the only real item that the AM2 gave us was DDR2 support and a reduction in the amount of cache that each CPU core had access to. Next came the Barcelona, the first native quad core CPU, and the issues around certain TLB errata that took AMD some time to fix (this fix included a performance hindering patch) this issue also leaked over into the Phenom line and resulted in one of the worst launches that AMD has ever had. AMD attempted to rally with a launch of a tri-core CPU making the claim that this was a separate design and not a quad with a failed core. Unfortunately it was discovered that this claim was not true, then it was also found that since these tri-cores were quads with a failed core they also suffered from the same TLB errata. AMD looks to be in a very bad way. However there appears to be light at the end of the long dark tunnel that AMD is in. This comes in the form of the newly released B3 stepping of both the Phenom X4 (Quad Core) and now the Phenom X3 (Tri Core). Will this latest release give us the performance that we want and used to expect from AMD? Or is this another step down the tunnel into darkness? We have a Phenom X3 8750 in the lab and are going to find out.

Product: AMD Phenom X3 (8750)      
Author: Sean Kalinich
Reviewed on: April 18^th 2008
Product cost: $195.00
Manufacturer: AMD
Spelling and Grammatical editor: Jason Hambly, Paul Mercer

{mospagebreak title=Why Three Cores?}
Why Three Cores?

“…yes, but this one goes to 11…”
Many people have asked and indeed are still asking why three cores? What is wrong with just producing a Dual and a Quad line? These questions are further enhanced by the knowledge that the X3 is nothing more than an X4 with a single core disabled due to it failing QA at the FAB. One would wonder why you would want to sell your failures. AMD’s response to this is that they want to be able to provide a balanced line up for consumers. This gives them a line up with all cores numbers from two to four. But can the price/performance make the X3 worth the purchase? What market is this pointed at? Again AMD has a ready answer; this CPU is aimed at the market and environment right between the dual and quad segments. This is the mid-range user, the casual gamer and the HTPC enthusiast.
Here is an example:
Say you are looking to build an HTPC that you want to decode BluRay DVD on( for playback), can play a few games on and still have enough power and head room to encode your old DVDs to AVI without taking a couple of years. Most of us would not drop a $600 quad into a system like this nor would we grab up an 8800GTX for use in this system. But we might put in a dual core CPU only to find out that the average software driven BluRay player (CyberLink Power DVD) brings our inexpensive dual core to its knees. And that many games are looking for more power (on a UMA GPU setup) than the average dual core can offer. Here is where the X3 enters the game, it is a low cost CPU (the 2.4GHz X3 runs for $195.00) and have more power behind it than the same speed dual core CPU. If you drop this into a good UMA mainboard (see our evaluation of the GIGABYTE GA-MA78GM-S2H), 4GB of RAM and a 250GB HDD you now have a very good HTPC system that will also give you good gaming performance (at 1024×768 to 1280×1024 resolutions). The X3 will also fit into the mid range casual gamers market in a similar situation you would be able to pick up the X3 for right around the same cost as a dual core and for less than a quad and still get enough power to run the games and apps you want.
With all this in mind the concept of the X3 makes sense but how does it perform in practice?

{mospagebreak title=It’s Just So Odd}
Trouble in Paradise?:
“you know that every silver lining has a grey cloud around it..”
Now there is something to bear in mind as you contemplate the purchase of the Phenom X3. That is the nature of multi-processing. As it stands now multi-processing is still setup to be symmetric; what this means is that most systems are going to look for extra CPUs in even pairs for SMP to work. Where does this leave the X3 and its odd-man-out core? Well it seems there are a few fixes out to cover this small (yet glaring) issue. Microsoft has released a hot fix for 32bit Vista and Server 2008 (http://support.microsoft.com/kb/950182/en-us) but not for the x64 version. This is a nice fix and appears to help with certain applications but it should be an update and not a fix that the end user has to go and get.
There are also fixes for other applications that experience the even core only problem, although I do not have a complete listing of them at the time. Still even with the fixes for the odd number of cores this small item will impact the sales and adoption of the X3 into the mainstream. Users that this CPU is aimed at will want to see instant performance. This is especially in light of the recent lackluster performance of AMD’s releases.

{mospagebreak title=Performance I}
Performance:
Usually I would go into the technical details of the CPU listing its various features and merits.
However the B3 stepping of the X3 is not a new CPU so all of the technical details have been talked about ad nausem. The only really new item is the correction of the TLB errata that caused so much heart ache in the original releases.

So with all of that out of the way let’s dive into performance: Prior to this evaluation I asked the readers of our forum what types of performance tests they wanted to see the Phenom X3 run through. As you can imagine it was a long list, I have selected a few out of the group and present them here for you.

 Test System (HTPC/Entry)
GIGABYTE GA-MA78GM-S2H mainboard
AMD Phenom X3 8750 2.4GHz
2GB Kingston KHX7200D2K2/2G
Western Digital 74GB 10k RPM Raptor HDD
Mushkin 1.1KW PSU
Sony 16x DVD-ROM
Microsoft Windows Vista Ultimate x64

The test system was put together with the HTPC/ entry level user in mind.
I will be following this up with the X3 on a 790FX board to see if it can handle high-end applications as well.

The tests:
I know you are eager to see what types of hoops I m going to make the X3 jump through.
The list is a little longer than I would normally use but I hope it will give a very good indication of the type of performance you can expect from the Phenom X3.

Futuremark:
What evaluation would be complete without one of Futuremark’s tests? Even though no synthetic test can ever replace real-world performance testing we all have a love for the number that we get at the end of running one of these tests (well those of us with high number love them).
I decided to use both 3DMark06 and PCMark Vantage in my testing as they both show different sides of CPU and system performance.
3DMark06:
My interest here is the CPU score, it looks like the X3 was able to gain a roughly 1000 point lead on the X2 5000+ despite being clocked 200MHz slower.

PCMark Vantage:
Again we see the performance advantage of the X3 and its extra core.

HyperPi 0.99b
This test another synthetic that spits out numbers at you (the number Pi calculated out to a ridiculous number of decimal places). It runs by spawning multiple instances of SuperPi Mod (from Xtremesystems.com) one for each core.  The reason I use this is to test the speed and stability of data transfer between the CPU, memory and HDD. If there are any issues between these three critical components you are going to see it. Either your number will be way off or the test will simply fail. HyperPi is also good for checking the performance of each core if you have a weak core it will run this calculation slower than the other cores.

The X3 also manages to outperform the X2 and shows more consistent performance across multiple cores.

Video Encoding:
Video encoding is something that is becoming more and more popular. Encoding applications are everywhere on the Internet now and with Hard Drive sizes reaching into the terabytes it is not surprising that many people are backing up their DVD collections to disk.  For my testing I used AutoGK to encode a 2 hour movie. I maintained the same video quality (100% quality) and audio format and encoded to Xvid AVI format.

Time in Hours, Minutes, and Seconds (lower is better)

The X3 managed to knock ten minutes off of the encoding time over the X2 5000+ This is primarily due to having the extra core in play.

Cinebench R10.0
Cinebench is a testing suite from Maxon (makers of Cinema4D) and is designed to test your systems ability to work with Cinema4D. It included two CPU based render tests (single core and multi-core) as well as an OpenGL test to determine if your GPU is powerful enough.

Here in Cinebench we see the performance pattern continue. The X3 8750 performs better on single core performance and, of course, in multi-core performance.

{mospagebreak title=Performance II}
Lightwave 9.0
Lightwave is not actually a benchmark but it a 3D animation software that allows you to create 3d animated models from the ground up and then to animate them. Lightwave has gone from being an Intel only product to an AMD optimized product and is now somewhere in the middle. With both sides seeing performance comparative to the actual speed of the CPU. For my testing I used the Space Station scene from the Lightwave 8.0 media set. My render was of frame 32 only with output set to 1080p HD resolution; AntiAliasing was performed using a PLD 5 Pass run with motion blur set to Gaussian sharp. Threads were set to 2 and 4 with segment memory at 128MB.

Two Threads:

Here was an interesting test. When I ran Lightwave with two threads on the X3 my times were only marginally faster than the 5000+, but when I changed the threads to four for the X3 the improvement was significantly better, I would not have expected that type of performance increase from just having a single extra core available.

Sisoft’s Sandra Pro:
Although not an indication of real world performance Sandra can provide you with a set of numbers that indicate your system’s raw performance.
For testing I only ran the CPU tests, including the Multi-Core efficiency test.

My results here were mixed; it is obvious that the X3 is a better performer and that this increase is due to the extra core, however when I ran the multi-core efficiency test the X3 8750 was only marginally better. I was surprised as I would have thought the new HT3.0 and Phenom design would have dramatically improved multi-core efficiency, but it would seem there was only a small improvement.

POVRay:
POVRay stands for Persistence of Vision RayTracer and is another 3D graphics application.
It also has a pretty nice render test that can utilize multiple cores. This test is a pretty good indication of the rendering power of your CPU.

No big surprises here the X3 out runs the X2 by quite a bit, again that extra core shows it value.

Passmark 6.1:
Passmark is sort of a combination between PCMark and Sandra. It allows for synthetic testing of your whole system or specific subsystems. I ran the CPU related tests to get a good feel for how the 8750 X3 would perform.

Photoshop CS3 filter testing:
Although Photoshop is not what you would call an entry level product, it is an industry standard one and for that reason can be used to give us a base line performance test to determine how well a CPU (or system) can handle image manipulation. For this test I ran the Driver Heaven (driverheaven.net) Photoshop test.

Time in Seconds (lower is Better)

The story is the same in Photoshop CS3, the X3 handily outperforms the X2 5000+

{mospagebreak title=Performance III}
Gaming Tests:
Gaming is going to rear its head in just about any evaluation. Even while testing server related hardware I have been asked if it could be used for gaming. The Phenom X3 is not meant as a top end CPU for gaming but is intended for use by the “casual gamer” not the hardcore 1337 gamers out there. But even with its target being the mid range gaming market the Phenom X3 needs to be able to handle the latest games. Someone might not need 1920×1200 resolution, but might still want to play Call of Duty 4 or might just want the extras found in Company of Heroes.

I used four fairly current titles and out of those I picked 2 First Person Shooters (one of which needed to be a DX10 game), one Third Person Driving game and a Real Time Strategy game. I felt this would cover the majority of game types that people would be interested in.

Call of Duty 4 Modern Warfare:
This is as the name would suggest the fourth installment in the Call of Duty series. As the name also suggests you are placed in a modern environment. During the game you play multiple rolls and go thought various covert and low-level combat situations. Game play can be almost nerve-wracking at times where you rush through the usual “bar-fight” AI that COD is so fond of. The graphics are a tad on the cartoonish side but that is overshadowed by the great story line and immersive game play. For my testing I ran through the Crew Expendable mission and measures the frame rates using FRAPS 2.9.4

In Call of Duty Modern Warfare the X3 did not beat out the X2-5000+ by much in terms of frame rate but it was a much smoother game with the X3 in than with the X2.

Bioshock:
This title from Looking Glass Studios is great game and follows in the tradition as SystemShock and SystemShock II where you play the unwitting pawn of the bad guy, but redeem yourself in the end as you thwart their plans and destroy them. Graphic are very well done with a DX10 surfaces options that add a little more atmosphere to the game and workload on your GPU.
I ran the Welcome to Rapture level up to where you first see the Big Daddy and the Daughter.

The first thing you will notice here is the truly awful DX10 performance that the 780G offers. Although AMD uses the 780G to show an integrated UMA GPU with DX10 support the performance is not there to justify the claim. Granted there is Hybrid Graphics that I have yet to test out. This may be the saving grace of the 780G, as I mentioned that is yet to be seen. However there is still the fact that the X3 was able to eke out an additional 2 frames per second over the X2 and the game play went from choppy and unplayable to merely sluggish.

DiRT:

Colin McRae’s DiRT is an rather fun driving game that puts you in a fairly accurate simulation of a rally car race. The graphics and physics are well done as is the AI of the other drivers in the game. For testing I used Rally World to run a single track with three laps.

The 8750 X3 is the winner here again with a small FPS increase but more to the point in the overall playability of the game. With the X2 5000+ the game seemed to be slow and unresponsive; with the X3 there was a new spring to the game.

Company of Heroes:
Company of Heroes is a semi historical real time strategy game that puts you in command of a company during the World War II. You start off on the beaches of Normandy and fight your way inland. The graphics are surprisingly good and the enemy AI also well done. For my testing I ran the Normandy level from after you get through the barbed wire at the shingle all the way to the destruction of all of the 88s.

One more time, the 8750 X3 takes the show. Showing the largest performance improvement of the four games I tested. Again not just in the smoothness of the frames but in overall responsiveness.

{mospagebreak title=Conclusion}
Conclusion:

I know that many of you will be wondering why I did not include a comparison with an Intel C2D. The honest truth is that I did not have time. I wanted to get this out to you on time and did not have a comparable board for Intel that was a close match for the UMA 780G board I was using. As soon as I have that board in, rest assured that I will be running these tests again. I also did not have time to explore the overclocking potential of the X3 or how well it performs on higher-end hardware. This too will also have to be a future article. The comparison that I did use was impressive none the less. The X3 consistently outperformed the higher clocked X2 5000+. Some of this will be due to the better memory controller in the Phenom and also the use of HT3 but for the most part the simple addition of a third core put the X3 way out in front. To gain almost 1000 point in PCMark is no easy task. Even 3dMark06 showed an impressive improvement in overall and CPU score. Photoshop testing also showed a healthy improvement indicating that the X3 has much more headroom for image manipulation. Looking at the results I have in hand here I can say that the new B3 stepping is an improvement and a step in the right direction. I was even impressed at how well the three cores were put to use, I was expecting to see a number of these tests not even touch the third core but all of them used it and all showed an improvement.  This opens up an interesting market to the mi range and value market. The Phenom X3 showed us that it has enough raw horse power to run applications like Lightwave at an acceptable level. How many students might consider this CPU over a more expensive Intel system to get the performance they need out of the CPU so that they have the extra money to spend on RAM, HDDs and Video cards? AMD has given us something new to think about there.  
AMD might not have completely plugged the leak in their performance boat but they have seriously stemmed the flow with this new stepping of the Phenom X3 and possibly opened up a new midrange market.

Discuss this review here

Acknowledgments:
I would like to thank folks at AMD (especially Damon Muzny) for their help in getting this evaluation put together

Introduction:
An Introduction is usually intended to introduce (hence the name) something to a reader. In bocks the introduction usually brings the hero into the picture and often gives us a glimpse of the villain.
In reviews and evaluations the Introduction is used as a means of introducing a product or manufacturer. Since Intel really needs no introduction and the QX9650 (Yorkfield) has been talked about at length I think that we can safely skip both of these requirements and move directly into what you really want to know.

How fast is Intel’s newest Extreme Quad Core CPU?  

{mospagebreak title=What’s New}
What’s new?
Besides the obvious 45nm die shrink and the 12MB of cache the Yorkfield has a few new tricks up its sleeve.
High-K Metal Gate:
This nice little trick enables Intel to reduce the amount of power needed to switch the transistors from one state to the other while reducing the amount of power leaked. It also reduces the amount of oxide leakage that CPUs traditionally leak during operation.
Next on the list is SSE4. This latest edition to Intel’s Streaming SIMD Extensions promises even more performance gains with the addition of support for two different vectored 32-bit integer
Multiply operations, 8-bit unsigned min/max operations and 16-bit and 32-bit signed and unsigned versions. All of this goes into improved performance in video encoding, 3D content creation, and streaming load instructions.
As mentioned about the Penryn will have a much larger version of the Advanced Smart Cache seen on the previous versions jumping up from 8MB to 12MB. They have also included an enhanced cache line split loads capability. This helps when data is on two separate lines in cache, what happens is that the split load function speculatively dispatches the split lines of Data ahead of other loads to speed up the read  of the split lines of data.
The Penryn will also see improvements in Virtual Machine performance.
Now this list above is not all that is new in the Penryn, but it gives you an idea that this new line is more than a simple refresh.

{mospagebreak title=Power}
Power:
Intel CPUs have traditionally been more power hungry than their AMD counterparts. However with the introduction of Hi-k and 45nm processes the new line should offer a better performance per watt point.
The QX9650 that was sent out defaults to 1.25v at stock speeds, this is a very decent reduction over the 1.35v of the 6850

{mospagebreak title=Performance}
Performance:
This is what most people read a review for, the performance numbers. So with that in mind we plugged the Penryn into a GA-P35-DS4 and threw our usual battery of tests. The results were very telling indeed.

3DMark06:
Here it is for all of you numbers people. I included both the total score and the CPU score to show the difference in performance of QX9650 when handling AI and other CPU based calculations.

As you can see while the CPU showed a 10-12% improvement the overall improvement in 3DMark06 was roughly 3%. This goes along with what I have been saying about gaming lately. Unless the game developers pull a rabbit out of their hat games will remain GPU bound for a long time.

PCMark Vantage:
Another Synthetic from FutureMark this suite does seem to give a good overall performance picture of a given system. But as with all synthetic testing it cannot really show what you will see when a system is put to actual use.

Here we begin to see the advantages of the 12MB of Smart Cache and SSE4 in 64 bit. While the usual performance lead was 6% when running the 64 bit version of Vantage the lead widens to just over 10%.
This is all at the same clock speed.

SuperPi:
Another test that is used by the numbers crowd, SuperPi, is a good test to see how well your CPU, Memory, and HDD perform together.

The Yorkfiled, with its 12MB of Cache performs exceptionally well since many of the calculations can be done right in Cache.

Video Encoding:
For my encoding tests I used AutoGK version 4.5 and a Home Made DVD with video clips I have shot over the years.

SSE4 rears its head again, as we see quite a nice drop in encoding times.

CineBench:
CineBench is a 3D content creation benchmark that tests your systems ability to handle 3d animation.
For my testing here I used the two CPU tests; one is designed to test a single CPU or Core in your system and the other for Multiple Cores or CPUs.

With its extra Cache and SSE4 the Yorkfiled walks away with this test as well.

POVRay:
POVRay is another 3D Graphics application that has its own built in Benchmark for testing your system.
This includes a stock graphic for rendering and can be run on one CPU or all CPUs.
For my testing here I ran the benchmark on all CPU cores.

Not much to say here but the Yorkfiled wins again.

{mospagebreak title=Performance Gaming}
Gaming:
The games below were all chosen because they are supposed to be Multi CPU/Core aware and can take advantage of them.

BIOShock:
A great game from Through the Looking Glass studios, these are the same people that brought us System Shock and the Thief series.
The graphics in BIOShock are a little cartoonish (Despite being DX10) and the enemies, although fast and hard to hit, are not very smart. It is still a very enjoyable game.

Time Shift:
Time Shift is a fun game where you are immersed in a struggle to restore future history to what it should be.
Here is another game with very cartoonish graphics but unlike BIOShock the enemies are more challenging and work in teams to kill you.

HalfLife 2 Portal:
Portal is one of the most amazing short games to come out. This is not surprising since it is a brain child of the people over at Valve. Ever since the first HalfLife hit and put you into the Hazard Suit of Dr Gordon Freeman Valve has been putting out one hit after another.
In Portal there is no real enemy AI but a significant amount of physics, particle effects, and of course the portals themselves. These items are real time tunnels to different areas on the map and when using a map editor are even able to link between maps.

Crysis (Demo):
Brought to you by the same people that wrote FarCry, Crysis is a visually immersive game with many amazing graphical effects. Lurking behind all of this is a monster of an AI engine and one of the most all encompassing environment mapping systems.
What this means is that you can interact with just about everything in the game; from shooting out a tire on a moving vehicle to blowing up the gas can strapped to its back. Your shots have different effects on different surfaces or different parts of your enemy’s bodies.  Although the environment mapping is present and operational in the Demo, the AI was a disappointment I imagine that the full game does this better.

What can I say, the performance of the Penryn was roughly 20% better than the Yorkfield this is a pretty nice jump when you consider it is the same clock speed and has the same FSB.
The one place, as many of you have come to find out, where you are not really going to see performance increases is in gaming. Even with the resolution turned down there was very little performance difference. Until the game developers start writing games to truly take advantage of a Multi-Core CPU we can expect gaming to be GPU limited for some time.
This is even with games that brag about Multi-Core support like Crysis.
But in applications that require multi-threading and in multitasking the QX9650 rises above the rest.

{mospagebreak title=Price/Warranty}
Price/Warranty:
The pricing is about what you would expect from the top of the line Intel CPU and comes in at a hefty $1100.00 US it will make a serious dent in your wallet (and in some cases your savings account).

Intel offers a Three Year warranty on their retail CPUs.

{mospagebreak title=Conclusion}
Conclusion:
What can I say that has not already been said? The QX9650 is an amazing CPU and simply outperforms anything that AMD has on the market right now in terms of general performance and multi-media creation and playback.

Where gaming is concerned you are not going to see much of an increase in terms of “frames-per-second” based performance, what you will see is an increase in the ability of your system (equipped with the QX9650) to handle all of the AI and other streaming data that is not GPU bound.

This will free up your system to handle the massive amounts of data being processed by the GPU.

Will this make your frame rate go up? No, but it will make the game feel more fluid and responsive. In games with heavy AI the enemies will seem to react and respond faster this is especially apparent on the higher difficulty levels.

Overclocking on the QX9650 was very easy, without much fuss at all I was able to get the QX9650 to run at 1GHz over stock speeds with no RAM tweaks and only a minor bump in voltage.

Intel also walks away with the Multi-Tasking crown despite the limitations of the non-shared Cache between the two sets of cores. The 12MB of cache tends to make up for this lack and the higher speed RAM available more than removes that perceived bottleneck.

Even a few months after its release, the QX9650 still stands out as the performance leader in an increasingly one company race.

 
Product: Core Extreme Quad QX6850
Author: Sean Kalinich
Category: Enthusiast
Reviewed on: July 16th 2007
Product cost: TBA
Manufacturer: Intel
Spelling and Grammatical editor:  Carlos Echenique, Paul Mercer

Performance:
Although this is not a full blown review of the QX6850 (just like the E6750 that will have to come later) we still wanted to give you a good look into what the QX6850 can do.  I tested this against the QX6700 both CPUs running at stock speeds. I did not include an AMD CPU as I have not been able t get a quad-core sample from them for comparison. Again, as with the E6750 CPU preview, I broke the testing into two parts. Synthetic and Real-World gaming.

The results of my testing are listed below.
The test system was:

Synthetics:
Ah synthetic testing, always a source of controversy and bickering. Is a synthetic test a real gauge of what a system can do? I personally say no, as synthetic tests can be fooled by certain optimizations in drivers (anyone remember the quake III testing fiasco?) and they really do not take the whole system into account. I do include them as they can be a good reference point to how a system should run with similar hardware to the test system. My tests of choice are below.

Bear in mind as you look at these numbers. The QX6850 not only has a faster FSB but a fast clock speed. I did not want to get into over clocking the QX6700 to match the speed as overclocking a CPU for testing leads to other errors in comparison.

3DMark06

Although 3DMatk06 is aimed at gaming performance it does have a fairly good test of CPU performance on a game level. The average gain for the overall test was only 4.5% while the average CPU test gain was 21.71%.

PCMark05

PCMark05 also shows an over all improvement, this time it is 18.53%. I am still not convinced that this is not just the improvement in overall speed (2.66GHz vs 3.0GHz).

POVRay 3.7 Beta

POVRay 3.7 Beta is a fairly new test but has shown itself to be an honest representation of CPU based rendering. Here the performance seems to be more than just the jump to 3.0GHz. The improvement is a nice 21.26%

CineBench 9.5

Cinebench 9.5 is another render test, although they have more than just CPU based test. CineBench has both single and multi-core tests. Here the single core improvement is 20.99% and Multi-Core is 19.60%. This shows some of the overhead associated with multiple cores, although the actual render scores show more plainly the benefits of Multi-Core.

Video Encoding:

Now here is something interesting, at first I thought I had made a mistake. I ran the test over and over and the results were the same. It seems that the QX6700 is simply faster at Dixv encoding when using AutoGK.

SuperPi Mod:
QX6700 Multi-Core 32M Run (With Core Affinity)

QX6700 Single 32M Run

QX6700 Multi-Core 32M Run (Without Core Affinity)

QX6850 Multi-Core 32M Run (With Core Affinity)

QX6850 Single 32M Run

QX6850 Multi-Core 32M Run (Without Core Affinity)

All times in Minutes and Seconds 

QX6700 Multi-Core 1M Run (With Core Affinity)

QX6700 Single 1M Run

QX6700 Multi-Core 1M Run (Without Core Affinity)

QX6850 Multi-Core 1M Run (With Core Affinity)

QX6850 Single 1M Run

QX6850 Multi-Core 1M Run (Without Core Affinity)

Time in Seconds

Gaming:
It has long been known that, for gaming, the CPU plays very little into performance at high resolutions. That is why for all of my testing here I used a relatively low resolution. 1024×768. I turned off all of the eye candy as well. This will show the performance differences in how the CPU handles AI, CPU based physics, and other items that do not involve the GPU.  Again bear in mind the clock and FSB speed differences between the two CPUs used in the comparison.

Need for Speed Carbon:

With Need for Speed Carbon we see a performance gain, but again this is more than likely due to the actual CPU speed vs the FSB speed.

F.E.A.R.

F.E.A.R. shows no performance gain or loss.

Again no gain or loss worth noting.

HalfLife 2 Episode 1

Once more we see no gain or loss.

Conclusion:
So as we saw in the preview of the E6750 there is not much of an improvement over the E6700, but the QX6850 with the faster FSB and clock speed is an improvement over the currently available quad. This CPU now sits with the Multi-tasking performance crown squarely on its head. I hope that AMD can get back into the game and at least offer some competition for this CPU, but as AMD has always chosen the tight-lipped approach on new products we will not know until they are released. But with the market the way it is, many people will not want to jump ship on a new and untested CPU when they have a known power house they can pickup. Adding to this threat is the fact that Intel will have its very own native Quad-Core coming out soon and if the performance of the current non-native Quad-Core CPUs are any gauge they will be very fast indeed.

Acknowledgments:
Thanks goes out to Dan Snyder at Intel for the QX6850 (and the QX6700) used in this evaluation.

Intro:
A few days ago the net was hit with a barrage of performance reviews of Intel’s latest stop-gap
the E6750. This CPU on the surface seems to be nothing more than an E6420 with a 1333MHz FSB and a lower TDP. But with this refresh of the Conroe line (due to the announcement of Bearlake and the native 1333MHz FSB CPUs) comes the question “Will we see any real performance increase? “
We took the time to find out.

Product: Core 2 Duo E6750
Author: Sean Kalinich
Category: Enthusiast
Reviewed on: July 7th 2007
Product cost: TBA
Manufacturer: Intel
Spelling and Grammatical editor:  Carlos Echenique, Paul Mercer

Performance:
Although this is not a full blown review of the E6750 (that will have to come later) we still wanted to give you a good look into what this newest refresh can do.  Unfortunately, I did not have an E6700 handy for comparison. I did have a QX6700 handy and pitted the E6750 against this with two cores disabled( although I had no way to confirm this I tried to make sure that both cores were using the same 4MB of cache, as the Kentsfield cannot share cache between CPU dies).. Now you will notice that there are no AMD CPUs used in this eval.That is due to a complete lack of any of the newer CPU types being available from AMD. I if you would like to see the performance numbers from the last AMD CPU we had in for testing check out our QX6700 and Core 2 Duo reviews.Testing was broken down into two main groups, synthetic tests and gaming performance. Over Clocking will be covered later.

The results of my testing are listed below.
The test system was:
Intel Core 2 Quad QX6700 (with cores two and three disabled)
Intel Core 2 Duo E6750
eVGA 680i SLI Mainboard
Foxconn FV-N88XMAD2-ON GeForce 8800GTX 768MB
Corsair CM2X1024 8500C5 (2GB running at 1066MHz)
2x Western Digital 76GB 10k RPM Raptors (Non RAID)
PC Power & Cooling Turbo Cool T1KE-4E 1KW PSU
Corsair Nautilus 500 external water cooler for CPU.

(a word on the eVGA 680i SLI mainboard, This is the original reference board sent out by nVidia. I have never truly been able to get it to run stable. I used it for this test only because it was available and with a BIOS update saw the E6750′s 1333FSB natively. Although I feel the times and scores here are accurate some margin for error should be allowed due to the instability of the platform.)

Synthetics:
3DMark06:

As you can see the E6750 performs slightly better in 3DMark06. This is due to the faster CPU and increased bus speed.

PCMark05:

PCMark05, in Vista x64, enjoyed a larger performance gain as many of the tests in this suite are heavily CPU/Memory related.

POVRay:

Here the E6750 come out on top in Vista x64 but the Qx6700 manages to beat it in Vista x86.
CineBench 9.5:

In Cinebench 9.5 the performance gain is very slim but still there.

SuperPi Mod:

Time in Minutes and Seconds

Time in Minutes and Seconds

Time in Seconds

Time in Seconds

Well with SuperPi there is not much difference. In some places the E6750 wins by a hair but in others the Q6700 takes it.

Divx encoding:
Adding this under the synthetic benchmarks may seem odd but since this was a static test with a test DVD I did not feel it would be a true real-world test.

I used Auto Gordian Knot for my encoding and ran a movie that I authored for  a company I did some work for as my test DVD.

Time in hours:minutes:seconds

Gaming:

All games were run at minimal graphical settings and at 1024×768. This was to remove the GPU from the performance of the system as much as possible.
Need for Speed Carbon:

Need For Speed Carbon was a trade off, but the performance difference was minimal in both cases.

F.E.A.R.:

Wow, here we see a whole 2 frame difference. I am beginning to feel there is no performance gain to be had for gaming, but there are two more tests to run.DiRT:

Well this test again shows no major performance gain to be had by jumping to the 1333MHz FSB CPU.

HalfLife 2 Episode 1

Last test…

Well except for the 10 frame difference in Vista x64 with HalfLife 2 Episode 1 there really is no gaming performance gain to be found.

Conclusion (Sort of):

The numbers here do not show a huge difference. I would suspect that if you have an E6700 already you would not see a huge performance increase… However, that is not to say the E6750 is not a good move for Intel as where this piece of silicon shines is when you push it. We will be following up this performance preview with some overclocking results.  

{mospagebreak title=What?s New?}

In the end I ran the fan at 70% for my testing.

Now it seems the Intel engineers have been
working overtime and done a complete rework of Intel?s flagship CPU. Codenamed
Conroe, this CPU will sell under the name Core 2 Duo and Core 2 Duo Extreme,
these will be the desktop performance king of Intel?s line up.

In fact for the last few months, there have been Conroe
reviews popping up like weeds on the Internet. These reviews that many have
been flaunting have been with pre-release, non-(Intel)supported engineering
samples that were ?leaked? from Intel (the rumors say from Intel TW). You can
read review after review that parade amazing numbers in front of your eyes and
generate just as amazing page hit numbers for the sites that published them.
The sad fact of most of these is that the CPUs they have been using are not
what the consumer will be getting when they shell out their hard earned money
for one. Although I have to marvel at the person that thought up this marketing
strategy. It is a brilliant one, flood the Internet sites with samples so that
the consumer will see amazing numbers from Conroe. This undermines the
confidence in AMD as the AM2 is not showing that kind of performance.

{mospagebreak title=Architecture}

Wide Dynamic Exectution: This allows for four full instructions to be executed simlutaneously. (As opposed to the three instructions the Mobile and Netburst architectures can handle)

Macrofusion: Macrofusion allows for similar instruction pairs to be combined into a single internal instruction (called a micro-op)

Improved Arithmetic Logic Unit: This assists in the macrofusion processing allowing for greater performance without an added cost in energy.

Memory Disambiguation: An advanced out-of -order instruction that allows the poccesor to intelligently load instructions from memory. While this sounds simple there is more to it. Normaly, a microprocessor cannot unload infomation stores to reschudle new loads due to possible memory dependancies. Memory Disambiguation allows the microprocessor to determine if there are any dependanices. If it finds that there are none, it can reschedule instructions to allow for faster and more efficient execution.

Advanced Digital Media Boost: Allows for execution of SSE instructions are a rate of one complete instruction per clock cycle, this is an imporvement over the previous one instruction per two clock cycles.

Intel
Core 2 Extreme X6800 (2.93GHz)/ Core 2 Duo E6700 (2.66GHz)
Intel Desktop Mainboard DX975XBX
Kingston HyperX DDR2-800 CL4 (SLI Ready)
Leadtek PX7900GTX TDH Extreme 512MB
2x Western Digital 10,000RPM Raptor SATA HDDs (RAID0)
ThermalTake Pure Power 680Watt PSU
SilverStone  NT06 CPU cooler
Generic 16x DVD-ROM
Generic 1.44MB Floppy
Microsoft Windows XP Professional/ Microsoft Windows XP x64 Edition

AMD
AM2 FX-62 (2.8GHz)/ X2 5000+ (2.6GHz)
MSI K9A Platinum
Corsair XMS2 Twin2X2048 6400C4 DDR2-800 CL4 (SLI Ready)
Leadtek PX7900GTX TDH Extreme 512MB
2x Western Digital 10,000RPM Raptor SATA HDDs (RAID0)
SilverStone Zeus ST75ZF 750 Watt PSU
Stock AM2 Cooler
Generic 16x DVD-ROM
Generic 1.44MB Floppy
Microsoft Windows XP Professional/ Microsoft Windows XP x64 Edition

For my AM2 system I went with an ATi chipset to remove the nVidia optimizations (SLI Ready RAM Link Boost etc). Since I do not have an nVidia based mainboard to test the Intel CPUs on I wanted to get as even a playing field as I could.

All games were run at 1024×768 with no
additional graphical options on. The only exception to this was HalfLife 2
Episode One where HDR was set to Full. The HDR used in HalfLife is a more CPU
intensive integer based HDR. These settings, for the most part, removed the GPU from the equation and should show how the CPU effects the basics of gameplay. We will be testing the Core 2 Line in real-world game play in a follow-up evaluation.
All testing was done during real gameplay (with reduced resolutions). Frame rates were recorded using FRAPS. 2.7.4

In Need for Speed the Core 2 line enjoys a rough 10% lead
across the board. With the E6700 outperforming even the FX-62.   

F.E.A.R.

Here we see Intel?s new monster running away with an average
40% performance lead.

Call of Duty 2:

Score another one for Intel as the FX-62 again
has trouble keeping up with even the E6700 

LightWave 3D:

The second in our rendering applications Newtek?s Lightwave
3D has been the choice of many professionals for years now. Starting with
version 8.5 Newtek included 64 bit support. In our article on the benefits of
64 bit over 32 bit found here we saw how a multiple core SMP system was able to
take advantage of the 64 bit memory addressing and run rings around a similar
32 bit system. With the Core 2 CPUs we have multiple cores and 4MB of cache to
buffer rendering information add into all of this, the EM64T instructions.

Time in Hours, Minutes, and Seconds (lower is better)

PhotoShop CS 2:

SuperPi 32M Single Run

Time in Minutes and Seconds (lower is better)

SuperPi 1M Single Run

Time in Seconds (lower is better)

32M Dual run with core affinity set

Time in Minutes and Seconds (lower is better)

32M Run No affinity

Time in Minutes and Seconds (lower is better)

Dual 1M with core affinity set

Time in Seconds (lower is better)

Dual 1M Run no affinity

Time in Seconds (lower is better)

{mospagebreak title=Numbers}

Numbers Game:

This section is for those of you that just HAVE to know how
well it does in PCMark05 and 3DMark06?. Oh and wait what does it do in Sandra?

Well here are all of the numbers to make you happy. 

	Intel X6800		AMD FX-62
	Windows XP Pro	Windows XP x64	Windows XP Pro	Windows XP x64
CPU Arithmetic
Multi-Media
Mem Bandwidth

Pricing for the Core 2 lineup will be as follows (in 1000 unit lots):

In all, as of this writing Intel?s new Core 2 CPUs are
the predator in the CPU jungle. Running down and devouring the current
competition.

So now you know the about the raw horse power of the new Core 2 line, how will it fare in real world usage and at real world resolutions? Check back for our follow on reviews where we will run through those topics and to tell you how Core 2 performs in the real world. 

Acknowledgements:
I would like to thank Dan Snyder at Intel for providing the Core 2 Duo, Core 2 Extreme and DX975XBX mainboard used in this evaluation.

Discuss this evaluation here

Introduction:

Now AMD has made another move, the move to the newer DDR2
standard. This evolutionary move is the AM2 CPU. AMD has been nice enough to
send us one for evaluation. Let?s take it out on the track for a few laps.

{mospagebreak title=Architecture}

Architecture:
220, 221 what ever it takes?  

Most of you are familiar with AMD?s architecture. For those
of you who are not, you can read up on it here on AMD?s website.

Specifications:

COMMON TRAITS FOR AM2 PROCESSORS:

AMD ATHLON 64 FX-62 PROCESSOR TECH SPECS:

AMD ATHLON 64 X2 5000+ PROCESSOR TECH SPECS:

DETAILS ON OTHER AM2 PROCESSORS:

Athlon 64 and Sempron:

 
{mospagebreak title=A New CPU?}

?Modular? Design:
I think the couch would look better over by the window?
 
As I mentioned in my rambling introduction, AMD has great
forethought in the way they design their CPUs. With the Athlon64, Turion, and
Opteron line it was planning for Multi-Core. They were designed from the
beginning to be multi-core; the same can be said for the rest of the design. If
you look at a picture of an AMD CPU you will notice that the memory controller
is offset just a little, this allows for an easy swap of this part of the CPU. Instead of a complete redesign to give you access to new memory, AMD can remove
the old and snap in the new.

Evolutionary Move:
Many of you are thinking of the AM2 as a new advance in
technology, while in many aspects this is correct. What it is not, is a leap
forward. Instead, the AM2 is an evolution of the Athlon 64; this is similar to
when AMD moved the Athlon Slot-A to Socket-A, or when they moved the Thunderbird
cores to use PC133 memory. Before this point DDR2 speeds and latencies were not
where they needed to be. Now with DDR2 800 and DDR2 1066 CL3 and 4 out and
readily available, the move makes more sense.
AMD can leverage the increased bandwidth of DDR2 and the
Athlon AM2?s on-die memory controller to provide you with a better platform.
AMD also moved to a 940 Pin Socket for the A64, although you
will not be able to put a 940-pin Opteron in the AM2 Socket, AMD has changed
the pin out to prevent the enterprising and outright stupid from attempting
this.

Power:
Power is everything in PCs. If you build a system that is
power starved you will have an unstable system, yet with the increasing power
demands of today?s hardware comes increased heat. To strike a balance, AMD has
been doing some real work on reducing the power needed to push their CPUs.
Announced, in conjunction with the AM2 release, is a series of Energy
Efficient CPUs that operate at 65 watts max.  Certain CPUs will be available in
35 watt max models. This does not mean that they run around 65 watts unless you
push them hard, this means that if you have them running flat out you might top
out near 65 watts.
The AM2 X2 5000+ that we tested was not an energy efficient
model, yet even so it still runs at a maximum of 89 Watts. AMD has said that
all future X2 CPUs will run at this wattage. With the rising cost of
electricity, the improvements in performance-per-watt offered by the AM2 are a
welcome change.
The Energy Efficient Models are listed below:

Energy Efficient models available at 65 Watts max power:

Energy Efficient models available at 35 Watts max power:

Heat:
Yeah but it?s a dry heat?

As mentioned above, with increased power you have increased
heat, the opposite is true for reducing your power. With the lower wattage CPUs,
AMD has produced a CPU that does not double as a small furnace. These CPUs run comparatively
cool and as can be seen below, run cooler than their older siblings run. The
X2 5000+ that we are testing here today runs off of only 1.3v.  That is around
.05v less than the 4800+ and is 200MHz faster by clock speed. The AM2
maintained a cool 38c under load in our testing.  When we added water cooling, it
stayed below 34c. Ambient temperature was 23c. By comparison, the Opteron 170
at stock, on air cooling tops out, under load, at 44c and the XE840 comes in at
an egg frying 72c under load, using the stock cooler, and 64c with water cooling. For cooling, AMD sent the same type of cooler that comes with the current FX series CPUs. For those of you with third party cooling solutions there is good news and bad, if your hold down bracket is single tab-per-side, it should work. The three tab-per-side brakets will not. AMD has also changed the placement of the screws on the mounting bracket. the current two screw solutions will not work as there are now four screws that hold the mounting bracket to the board.

Hardware Virtualization:
This is a huge leap forward, for those of us that run
multiple operating systems or who test software and would like to limit the
number of actual computers we work on, hardware virtualization is a godsend.
This technology allows you to run virtual operating systems at the hardware
level instead of through a software emulated layer.
Although the current virtual OS applications do not support
this feature, it will happen. The days of losing performance and productivity
because the emulated hardware could not keep up appear to be a thing of
the past.

{mospagebreak title=Test Systems and Methods}

Performance:
This is what you all want to know about, how well this
CPU will perform.  Will it be worth the money?  Should I start taking lunch money from
the other kids for this?  ?well maybe not that. I know that with all of the talk
and rumors around the Internet, performance is what you want to know about.

Methodology:
For this evaluation I wanted to test the CPUs with as many
real world tests as I could.
There will be very few synthetic benchmarks and no time
demos in this evaluation. They do not serve to show off real performance.  For comparison I have chosen a competing CPU
as well as an older CPU from AMD. I would have liked to run these tests against
the new Conroe CPUs, however at the time of writing, no samples were available.

For testing I used the fastest Intel CPU available to me; the Extreme Edition 840
which was run overclocked to give it the faster 1066 FSB and ran at 3.72 GHz
(266×14). For my AMD choice I used an Opteron 170 that runs smoothly at 2.6GHz.

AMD test platforms:
AMD AM2 X2 5000+
Asus M2N32-SLI Deluxe

1GB Corsair XMS2 DDR2-1066 (2×512) at DDR2-800 4-4-4-12

2x Leadtek PX7900GTX TDH 512MB Extreme (in Single and SLI for
game testing)
2x 74GB Western Digital Raptor 10,000 RPM SATA 150 HDDs
(RAID0)

Sony 16x DVD-ROM
 

SilverStone ST75ZF Zeus 750Watt Quad Rail PSU
Microsoft Windows XP Professional and Windows XP x64 Edition

Opteron 170 (clocked to 2.6 GHz)

Asus A8N32-SLI Deluxe
1GB GeIL One DDR 400 @217MHz 
2-3-3-5

2x Leadtek PX7900GTX TDH 512MB Extreme (in Single and SLI for
game testing)
2x 74GB Western Digital Raptor 10,000 RPM SATA 150 HDDs
(RAID0)

Sony 16x DVD-ROM

SilverStone ST75ZF Zeus 750Watt Quad Rail PSU

Microsoft Windows XP Professional and Windows XP x64 Edition

Intel Platform:
Intel Pentium Extreme Edition 840 @3.72 GHZ (266×14) Hyper-Threading
enabled
Asus P5WD2-E Premium
1GB Corsair XMS2 DDR2-1066 (2×512) at DDR2-1066 5-5-5-15 2T

Leadtek PX7900GTX TDH 512MB Extreme

2x 74GB Western Digital Raptor 10,000 RPM SATA 150 HDDs
(RAID0)

Sony 16x DVD-ROM

SilverStone ST75ZF Zeus 750Watt Quad Rail PSU

Microsoft Windows XP Professional and Windows XP x64 Edition

My Choice of Intel platform was due to an unfortunate
occurrence.
My two Intel SLI boards were unavailable. One died a tragic
death due to a bad PSU and the other was on loan to another site for testing.
This left only the Asus P5WD2-E Premium as an acceptable mainboard for
comparison. Unfortunately, I was unable to conduct any SLI tests on this board
and only single card scores are shown below.
Once either of the two Intel SLI boards that we have in-house
is available, I will add in the SLI scores from that board.
Settings for the 7900GTX, unless otherwise noted, were
global profile for all games with Transparency AntiAliasing set to Supersample.
In addition to the software used for testing I also wanted
to simulate a real world environment, for this I installed the following
commonly used applications:
Microsoft Office 2003

SpeedFan
Diskeeper 10

Avast Home Edition

Acrobat Reader 
A HighSpeed PC TechStation was used for all testing.

Corsair?s Nautilus500 was used for cooling the overclocked
Opteron and the Intel XE840.

{mospagebreak title=Performance Tests}

The Tests:

Gaming: These four game titles have heavy physics and particle
effects throughout the game. They are in many cases capable of bringing even
the fastest system to a crawl.

Need For Speed Most Wanted: Although this is an older title, the physics involved with
high speed driving that the game emulates can have a serious impact on
performance and frame rate. Add into all of this physics fun, multiple cars
with their own physics, lighting effects, etc., and you not only have a
visually impressive game, but also one that can put a hurting on your CPU. For
testing I ran the Challenge test number Nine, this track is very winding with
many scenery changes, and in many cases the weather changes during the run
altering the lighting and shadows.

Windows XP Pro	Windows XP x64

Here, as in most games, the AMD AM2 is simply the faster
CPU. The feel of the game was smoother and more responsive on the AM2.
Considering the AM2 was run stock against two overclocked CPUs and still kept
around a 5 frame per second advantage.

F.E.A.R.
If you have not heard of this game you need to put
down the Funions and get out of the house. F.E.A.R. is a Monolith Productions
Game. Monolith has an inclination towards the darker side of gaming with titles
like Blood and Blood 2.  They incorporate a supernatural and creepy feel into an
immersive gaming experience. F.E.A.R. combines a good story line, impressive physics,
startlingly intelligent AI and heavy particle effects into a great game. For my
F.E.A.R. testing I sampled frame rates in the Interval -02 – Insertion level
from the entry at the gate to the point where you are blown out of the window
by the little girl. This section has quite a few combat sequences with some
good AI routines. The enemy clone soldiers perform complex fire and maneuver
tactics. Visual setting were 1600×1200 with all other settings at Maximum,
Volumetric Lighting and soft shadows were also enabled, Computer settings were
at Maximum and Difficulty was set to Extreme , this allowed from more complex AI routines
and more cunning enemy movements

Windows XP Pro	Windows XP x64

Our F.E.A.R testing again reveals the AM2?s ability to
handle complex gaming instructions faster.

Call of Duty 2:
Another game with quit a bit of action, Call of Duty 2, is a
bar fight of a game, with enemies around every corner and allies that jump
right in front of you while you?re shooting. The scenery around you is quite
well done, with many smoke and particle effects, this game can put a serious
dent in a system with a weak CPU. Again for testing, I chose settings that I
feel would impact the CPU the most. 1600×1200 with everything set to its
highest level and difficulty set on Veteran.
The level used for testing was the Demolition
portion of the Winter War level. This was played from start to the destruction of the German building.

Windows XP Pro	Windows XP x64

Call of Duty 2 shows what the other game tests have already
shown, even against an overclocked CPU the AM2 is able to maintain its lead.

Oblivion:
ElderScrolls IV Oblivion is a game that everyone is talking
about (and not all of it good). This game has a way of just plain putting a
hurting on your system. Both ATi and nVidia have had to patch their drivers
just for this game. There is also a heavy toll on your CPU as the game seems to
stutter and drag on slower CPUs. Memory is also important, and either not
enough or not fast enough memory, will impact your gameplay.
The display settings for Oblivion were 1600×1200, with all
other detail settings on their highest setting. AA was disabled as Oblivion
does not support HDR and AA at the same time.
The testing level I used was the opening level where you are
attempting to escape through the Emperor?s escape route. I followed this path
until I reached the second wooden door.

Oblivion was a pain to test; there is no true support for XP
x64 or the 7900GTX, so the auto hardware detection would not work properly,
requiring the settings to be manually set. In addition, because there is no
support for the 7900GTX I had to manually enable the SLI function. Without this I
saw no benefit from SLI at all. Once I had the settings correct, the game had
an annoying habit of cutting to a blank screen. The menus would work but there
was no display. I was able to tweak some of the settings, and setting the
renderinfo.txt to ?read only? helped with this. Once I had this problem out of
the way
I was able to complete my testing.

Windows XP Pro	Windows XP x64

Again, as in the other tests, the AM2 has a slight advantage
in speed against the overclocked Opteron and XE840.

{mospagebreak title=Application Tests}
Cinebench:

A new test at Planetx64 Cinebench is a nice tool that can be
used to gauge the rendering power of your system. Cinebench is a product of
Maxon, the same people who developed Cinema 4D. According to Maxon, Cinebench
is a real-world test for rendering power and is able to accurately show off a
particular system?s ability render in real-time. Cinebench is especially CPU
intensive, and is able to show scaling from one to up to 16 CPUs. I chose this
test to show off not only the raw rendering power of the AMD AM2 CPUs but also
the benefit of the x64 instructions built into all AMD Athlon 64s. (See our Lightwave 3D 8.5
review for more information.)

Here, the raw power of the AMD CPUs are evident, easily outperforming the
overclocked XE840 in the rendering tests. The AM2 still keeps its head above
the overclocked Opteron 170, even with the smaller cache size.

SuperPi:
SuperPi is a test that just about every PC enthusiast and
overclocker knows about and uses. This test relies heavily on your CPU and
CPU-to-memory speeds to calculate Pi out to some impressive decimal places. For
these tests, since all of the CPUs tested are Dual Core, I ran two tests
simultaneously. This shows the stability between the cores.

	XE840 @3.72		Opteron 170 @2.6		AM2 X2 5000+
	Core 0	Core 1	Core 0	Core 1	Core 0	Core 1
Windows XP Pro	29:29.046	29:33.234	29:45.348	30.54.297	31:37.127	31:24.419&
Windows XP x64	38:50.734	38:32.766	31:45.078	31:45.969	30:36.828	30:33.735

SuperPi Times in Minutes and Seconds Lower is Better

The scores here show the benefits a larger CPU cache can
have. With the 1066 FSB, the XE840 is able to outperform the AM2 and the Opteron
170. The XE840 has two megs of cache, one meg for each CPU core, just as the
Opteron 170 does. The relatively small instructions for the SuperPi
calculations are easily stored in cache for execution.

However it still requires a 500MHz OC on the XE840, bringing
it to 1.1GHz faster than the AM2, to achieve this win. An interesting issue arose in our 64-bit testing on the XE840. It would seem that the XE840 is not
as nimble in XP x64 when running SuperPi.

Easy DVD to DivX SVCD converter:
For DivX encoding, I used Easy DVD to DivX SVCD converter,
this application is capable of using multiple CODECs for both audio and video
encoding. DivX 6.2.2 with CD Quality Audio was used for all encoding.

	XE840 @3.72	Opteron 170 @2.6	AM2 X2 5000+
Windows Xp Pro	0:43:42	0:57:48	1:02:35
Windows XP x64	0:39:29	1:01:12	0:40:21

Again, in encoding, the larger cache combined with the
faster bus and memory speeds helps the Opteron and the XE840. The faster RAID
solution on the Intel i975x also helps the XE840. These score are only possible
when the XE840 is overclocked. At stock speeds the XE840 is not able to keep up
at all.

{mospagebreak title=Pricing}
Pricing:
The pricing on the new AM2s has been on the Internet for a
while now. Some of it is accurate, some is way off of the mark. I have included
below the pricing direct from AMD:

AM2 PRICING:

FX-62:	$1,031
X2 5000+:	$696
X2 4800+:	$645
X2 4600+:	$558
X2 4400+:	$470
X2 4200+:	$365
X2 4000+	$328
X2 3800+:	$303
Athlon 64 3800+:	$290
Athlon 64 3500+:	$189
Sempron 3600+:	$123
Sempron 3500+:	$109
Sempron 3400+:	$97
Sempron 3200+:	$87
Sempron 3000+:	$77

AM2 PRICING (ENERGY EFFICIENT MODELS):

65 WATTS:
X2 4800+	$671
X2 4600+	$601
X2 4400+	$514
X2 4200+	$417
X2 4000+	$353
X2 3800+	$323
35 WATTS:
X2 3800+	$364
Athlon 64 3500+	$231
Sempron 3400+	$145
Sempron 3200+	$119
Sempron 3000+	$101

{mospagebreak title=Conclusion}
Conclusion:
AMD has proven themselves again and again as a manufacturer
that can deliver innovative products, performance, and stability.  With the AM2 we see another indication that
AMD is not only thinking ahead, but also ensuring that they provide the
consumer a quality product. The decision to wait to implement DDR2 support
until the technology matured was a solid one. This is what we have come to
expect from AMD, good solid products.
By comparison Intel has chosen to address
the latency issues inherent in their designs, not by improving on the original
concept but by stuffing as much cache into the CPU as possible. It seems they
have also ignored the cost of adding this extra cache in terms of power and
heat. Intel says Conroe will be a departure from
this, but Conroe
is not here yet.

Is the AM2 a revolutionary new CPU that will amaze you? Not
really, if you are looking for that, prepare to be disappointed. What AM2
represents is an evolution of the Athlon 64 line. AMD has done quite a job,
reducing power requirements, heat output, and changing the way the CPU
performs. This evolution of the AM2 combined with new chipsets and improved
DDR2, IS a revolutionary change. The
performance that can be had from using the new AM2 platform is going to
impress.

In our testing we found that the AM2 X2 5000+ was able to
outperform our overclocked Opteron 170 and XE840; this is with ? the cache of
those two CPUs. This is no small feat. Generally when you compare two CPUs of
the same clock speed but different cache sizes, the one with more cache will
win. This was not the case in with the AM2. Also, DDR2 RAM still does not have
latencies that come close to DDR. The DDR2-800 we tested runs at 4-4-4-12 while
our DDR-400 ran at 2-3-3-5 and was capable of 1.5-3-3-7. So with less cache and
a higher latency the AM2 X2 5000+ we tested still was faster than our
overclocked CPU by a small margin. The difference between the XE840 and the AM2
X2 5000+ was even more apparent.
With static instructions to work with (such as
in SuperPi and DivX encoding) the XE840 was able to leverage Netburst and the
larger cache to run faster. However, this was only in those two tests, and then
only with the overclock to the 1066 FSB. This is in no small part due to the
work AMD has put into the AM2, and it also ties in with the new platforms from
both nVidia and ATi.
Although at the time of writing, the new mainboards from ATi
were not available, they will be here soon. And I would not be surprised to see
the same increase in performance.

Should you upgrade to AM2? You will have to decide
that for yourself. I can say that with the across-the-board improvements I have
seen in my testing, I would say the AM2s will be well worth the upgrade. To put
it plainly, AMD has given us (the enthusiasts) another rock solid platform for
us to pound away on.

Acknowledgements:
I would like to thank Damon Muzny at AMD for providing the AM2 X2 5000+ and other hardware for this evaluation.

 

Discuss this evaluation here 

 

]]>