The AMD 690G/690V chipset was officially launched about a week ago. This is AMD's first integrated graphics platform and in our initial testing proved to be a serious competitor to the NVIDIA 6100 series and the Intel G965 in video performance. We have had a little extra time to work with the chipset and feel confident enough in the chipset and drivers to publish our first performance results today. While our initial results will concentrate on video, media, and audio application performance from a chipset level, we will be looking in-depth at each 690G board in our upcoming mATX roundup that features over ten boards from four different chipset suppliers.

We actually have had the MSI K9AGM2-FIH in our labs for several weeks and the initial performance results were somewhat disappointing based upon our expectations of the chipset (not the board). However, through the release of new BIOS code and driver updates from AMD, the board has turned into quite the performer in several areas. Of course the previous statement should be tempered to a certain degree as AMD's first IGP is competing against an NVIDIA chipset that has been out for close to 18 months and an Intel chipset that continues to disappoint us at just about every turn.

Our opinions about the basic performance level of current IGP solutions are not kind as we feel like the continued minimum functionality in such solutions creates issues with developers looking to move forward but who must always have a foot in the grave to ensure their products run on the lowest common denominator platform. To a certain degree, that platform has historically been Intel based as they are the world's largest graphics provider. This is a position that breeds frustration when the overall performance and feature sets are so low that one wonders why they even bother to update their IGP chipsets targeted to the home user.


Of course the introduction of Vista will ultimately benefit consumers and developers as it forces a certain base feature set and performance requirement for graphics hardware. However, even with DX9 functionality required for the full Vista UI, the performance and compatibility of current games with Intel hardware under Vista is dismal at best. All of this sounds like a lot of doom and gloom, and in some ways it is but such is the life of an integrated graphics platform in the hands of even the most casual gamer today. This is why we are always looking forward to seeing new integrated chipsets as competition drives innovation, no matter how small the improvements might be at times.

This is not to say that the AMD and NVIDIA solutions are that much better than the Intel G965; they are to a certain degree, but without earnest competition from Intel these solutions do just enough to stay ahead without actually encroaching on the discrete GPU market. However, at least these solutions provide a much higher degree of compatibility and performance with most games and applications. While running the latest games such as Oblivion or Supreme Commander require a resolution of 1024x768 or under with medium-low quality settings, at least a user has the chance to play the game until they can afford a better video solution although the experience will not be a pleasant one with the latest game releases.

Let's take a quick look at the specifications of the AMD 690G/690V chipsets and its performance against the other major IGP solutions.



Chipset Overview

The AMD 690G/690V chipset consists of an RS690 Northbridge and SB600 Southbridge. AMD's intent with this chipset is to provide an attractive alternative to the NVIDIA 6100 family, but more importantly they want to provide a total platform solution that is very competitive against the current Intel G965 family. The 690G is directed towards the consumer market with a heavy emphasis on multimedia capabilities via the X1250 graphics core, while the X1200 core on the 690V chipset is targeting the business market where AVIVO capabilities are not as important.


In the case of the X1250, it is no surprise that AMD has reached back to previous generation hardware for the base design of their new integrated GPU. Lower transistor counts mean smaller die sizes and lower costs, and the X100 series fits the bill with its lack of SM3.0 support and use of 24-bit floating point precision. The basic design for the X1250 is taken from the X700, with some modifications. While we would love to see Shader Model 3.0 support (which NVIDIA hardware has in their 6100 chipset and current Intel hardware claims to include), developers writing DX9 apps will still be designing for the SM2.0 target which the X1250 meets.

Many AVIVO features (including 10-bit per component processing) have been implemented on X1250, bringing higher quality video decoding to integrated graphics. Unfortunately, with this improvement comes some sacrifice, as the number of pipelines on the X1250 is cut down from the X700. The X1250 weighs in at 4 pixel shaders and like other X100 series hardware this also means 4 texture units, z-samples, and pixels per clock. The other major change when compared to the X700 is that the number of vertex shader units have gone from 6 to 0. All vertex shader operations are handled by the CPU.

The core clock speed operates at 400MHz and can be increased to 500MHz within the BIOS depending upon the board manufacturer. We have also overclocked one of our boards to 550MHz with a third party utility but performance unfortunately does not scale well in most games. We have seen performance improvements on average increase anywhere from 3%-12% percent depending upon the application.

As for memory, the GPU can handle up to 1 GB of memory, but support is once again dependent on the BIOS. AMD uses an optimized unified memory architecture (UMA) design, and all graphics memory is shared with system memory. For our tests, we found 256MB to be the sweet spot, as performance seemed to be degraded with 512MB or 1GB graphics memory, especially under Vista where the base memory requirements are significantly higher than XP. This may end up being different depending on implementation, but we will stick with the 256MB recommendation for now.

Looking beyond architecture, most people who will actually be using integrated graphics won't be bothered with games or high end 3D applications. This hardware will be most used for 2D and video applications. Let's take a look at the features we can expect in these areas.

Supporting a maximum resolution of 2560x1600, the X1250 can easily run any CRT at maximum resolution. This tops NVIDIA's 6150 max resolution of 1920x1440 and Intel's G965 at 2048x1536. As for output features, the video hardware supports S-Video, YPbPr, HDMI 1.3, and Dual-Link DVI. Of course, the actual interfaces available will depend on the implementation, but the HDMI and DVI ports will also support HDCP.

The GPU supports two independent display outputs, and both DVI and HDMI outputs can be used at the same time. The only caveat is that HDCP will only work over one digital output at a time. This isn't a huge issue, as most people won't be watching two different protected movies at the same time on a single computer. Also, in spite of the single display limitation, HDCP can be used over either HDMI or DVI. This gives the X1250 an advantage over graphics cards that initially supported HDCP. Many cards only allowed HDCP over one HDMI or DVI port while the other was always unprotected only.

As for HDMI, the audio support is enabled through an interface in the RS690 Northbridge while the SB600 Southbridge handles the HD audio controller interface. The standard HD audio codec is supplied by Realtek who has developed a driver package that allows the user to control both the HDMI and HD audio interfaces from a single application. The HDMI audio solution is capable of 32, 44.1 and 48kHz, 2 channel + AC3 (5.1) output.

For video acceleration features, the X1250 is capable of hardware acceleration of MPEG2 and WMV playback. MPEG4 playback decode is not hardware accelerated, but it is supported in software via the driver. DVD and TV (both SD and HD resolution) playback can be offloaded from the CPU, but we have seen some severe choppiness or blank screen issues with HD media formats at 1080p - although 720p worked fine. AMD has indicated that this issue will be addressed in a future driver and the chipset is fully capable of 1080p output with an upper end CPU and proper software support.

For those who wish to use discrete graphics alongside their integrated solution, AMD supports a feature they call SurroundView. This enables support for three independent monitors in systems with integrated and discrete AMD graphics. The feature works as advertised and may be useful for business users who want more than two monitors at a low price. Gamers who want more than two monitors will certainly have to take a different route.

The AMD 690G/690V utilizes the SB600 Southbridge that was introduced last May and continues to be a competitive offering, although both Intel and NVIDIA's latest chipsets are offering six SATA ports along with RAID 5 capability. However, LAN choices are left to the motherboard manufacturer's discretion. In general, the SB600 offers very good SATA and IDE performance while USB throughput slightly trails the Intel and NVIDIA offerings.



Test Setup

The Biostar TA690G AM2 was selected as our AMD 690G platform representative today. This choice does not indicate any perceivable performance differences that we have noticed when comparing it against the MSI K9AGM2 or ASUS M2A-VM boards; rather it was chosen after blindly pulling its name tag out of the lab hat. All three boards will be fully reviewed in our mATX roundup next week. At this time, each one offers basically the same base performance with features and pricing differentiating the boards from each other. The MSI feature list is very streamlined but still offers on-board HDMI output along with a price tag that should be south of $80. The ASUS and Biostar offerings are feature rich with price tags that will be around $85 and up. Performance and more importantly stability is excellent in all three boards.

Biostar TA690G AM2 / ASUS M2NPV-VM Testbed
Processor:	AMD Athlon 64 X2 5200+ Windsor Dual Core, 2.6GHz, 2 x 1MB L2 Cache, 13x Multiplier
CPU Voltage:	1.350V
Cooling:	Zalman 9500 Air Cooling
Power Supply:	Corsair CMPSU-620HX 620W
Memory:	OCZ Flex XLC PC2-6400 (2x1GB) (ProMOS Memory Chips)
Memory Settings:	3-4-4-9, 1.90V
Video Cards:	On-Board X1250, GeForce 6150
Video Drivers:	AMD 8.345, NVIDIA 15.00
Hard Drive:	Western Digital 74GB 10,000RPM SATA 16MB Buffer Seagate Barracuda 7200.10 500GB SATA 16MB Buffer
Optical Drives:	Plextor PX-760A, PX-B900A
Case:	Cooler Master CM Stacker 830
Operating System:	Windows Vista Home Premium 32-bit
.

MSI G965MDH Testbed
Processor:	Intel Core 2 Duo E6300 Dual Core, 1.86GHz, 2MB Unified Cache 1066FSB, 7x Multiplier
CPU Voltage:	1.3250V
Cooling:	Zalman 9500 Air Cooling
Power Supply:	Corsair CMPSU-620HX 620W
Memory:	OCZ Flex XLC PC2-6400 (2x1GB) (ProMOS Memory Chips)
Memory Settings:	4-4-4-12 1.85V
Video Cards:	On-board X3000
Video Drivers:	Intel 15.1
Hard Drive:	Western Digital 74GB 10,000RPM SATA 16MB Buffer Seagate Barracuda 7200.10 500GB SATA 16MB Buffer
Optical Drives:	Plextor PX-760A, PX-B900A
Case:	Cooler Master CM Stacker 830
Operating System:	Windows Vista Home Premium 32-bit
.

Our first decision was to decide what chipsets to test against. The natural comparison is the NVIDIA GeForce 6150 chipset that has ruled the AMD IGP market for well over a year now. We selected one of the better overall 6150 boards from a feature viewpoint, which is the ASUS M2NPV-VM. There was not enough time to compare the AMD 690G chipset to the current VIA and SIS AM2 offerings as these new products just arrived and we will look at them in the near future.

Our next decision and one that we typically do not do in our normal motherboard reviews was to test the board against a similar board offering but utilizing a different CPU manufacturer. AMD has been adamant that the combination of the 690G/690V chipset and AM2 processor is meant to provide an overall superior platform experience when compared against the Intel G965/Q965 with a Core 2 Duo processor. Of course, they also mean this when comparing their chipset against the NVIDIA products but at least with every NVIDIA AM2 chipset sold, there is a corresponding AMD processor requirement. In the end we chose the MSI G965MDH for our Intel G965 platform offering.

We selected the AMD Athlon 64 5200+ X2 and the Intel E6300 Core 2 Duo processors as our processor choices as both represent great bargains when comparing price against performance in their respective categories. We also switched to Microsoft Vista Home Premium 32-bit as our operating system of choice for this category. After speaking with several of the larger OEMs who are using this chipset and similar ones, we found out this OS choice will be the one most widely offered to consumers. It was a natural then that we would test on Vista Home Premium and a 2GB memory configuration as we suggest this amount as a minimum for Vista.

Our memory and hard drive choices are a little out of the norm for this category but since we are testing the chipset capabilities we were determined not to have a bottleneck in either area. We will test additional budget DDR2-800 memory from G.Skill and Wintec in our roundup along with comparison testing using an external video card. All other components in our test configurations are identical with the boards being set up in their default configurations except for memory settings being optimized to ensure maximum throughput on each board.

Overclocking, network, audio, and memory performance on individual boards will be covered in our mATX roundup. We will say that overclocking on the 690G platforms is in its infancy at this time as the initial BIOS releases concentrated on compatibility and base performance, not overclocking. We have received a performance oriented BIOS for the Biostar and ASUS boards that should allow overclocking now.

Our choice of software applications to test was based on programs that enjoy widespread use and produce repeatable and consistent results during testing. Microsoft Vista has thrown a monkey wrench into testing as the aggressive nature of the operating system to constantly optimize application loading and retrieval from memory or the storage system presented some interesting obstacles. This along with the lack of driver maturity will continue to present problems in the near future with benchmark selections. Our normal process was to change our power settings to performance, delete the contents of the Prefetch folder, and then reboot after each benchmark run. A lengthy process to be sure but one that resulted in consistency over the course of benchmark testing. All applications were run with administrator privileges.



Synthetic Graphics Performance

The 3DMark series of benchmarks developed and provided by Futuremark are among the most widely used tools for benchmark reporting and comparisons. Although the benchmarks are very useful for providing apple to apple comparisons across a broad array of GPU and CPU configurations they are not a substitute for actual application and gaming benchmarks. In this sense we consider the 3DMark benchmarks to be purely synthetic in nature but still valuable for providing consistent measurements of performance.


In our first test, the combination of the Intel Core 2 Duo and G965 makes for a great showing against the AM2 offerings. Okay, so we're being a bit sarcastic in that announcement as we consider these results to be anything but great. The Intel platform had no issues running the full 3DMark series but our AMD platforms could not complete the Shader Mark 3.0 tests. However, they exceeded the Intel platform scores in both the SM 2.0 and CPU tests. While the Intel platform passed the SM3.0 tests, this means little in actual game performance where the G965 failed to properly run games with SM3.0 capability.

General System Performance

The PCMark05 benchmark developed and provided by Futuremark was designed for determining overall system performance for the typical home computing user. This tool provides both system and component level benchmarking results utilizing subsets of real world applications or programs. This benchmark is useful for providing comparative results across a broad array of Graphics, CPU, Hard Disk, and Memory configurations along with multithreading results. In this sense we consider the PCMark benchmark to be both synthetic and real world in nature and it provides consistency in our benchmark results.


The margins are closer in the PCMark05 results with the 690G platform showing a minor advantage over the G965 and 6150 platforms. While this benchmark is designed around actual application usage, we will see if these results mirror our application testing.

Rendering Performance

We are using the Cinebench 9.5 benchmark as it tends to heavily stress the CPU subsystem while performing graphics modeling and rendering. Cinebench 9.5 features two different benchmarks with one test utilizing a single core and the second test showcasing the power of multiple cores in rendering the benchmark image. We utilize the standard multiple core benchmark demo and default settings.


The AM2 processors have always enjoyed an advantage in this test and the results continue to show the AM2 platform being dominant in this benchmark with a 26% advantage. The two AM2 platforms basically tie each other indicating CPU throughput is equal on either solution.



Media Encoding Performance

A traditional weak spot of the current AM2 processor offerings has been in the media encoding arena when compared to the Intel Core 2 Duo. Considering the main optimizations for the Core 2 Duo centered on overall media encoding performance, we were very interested in seeing how this platform compared to current AMD lineups. We are utilizing an updated video encoding test suite for this article that includes AnyDVD, Nero Recode 2, Windows Media Encoder 9, and QuickTime Pro 7.1.

Our first series of tests is quite easy - we take our original Office Space DVD and use AnyDVD and Nero Recode 2 to copy the full DVD to the hard drive without compression, thus providing an almost exact duplicate of the DVD. We then fired up Nero Recode 2, selected our Office Space copy on the hard drive, and performed a shrink operation to allow the entire movie along with extras to fit on a single 4.5GB DVD disc. We then go back and use Nero Recode to shrink the full DVD copy and burn it onto our DVD disc. We left all options on their defaults except we checked off the advanced analysis option. The scores reported include the full encoding process and are represented in minutes/seconds, with lower numbers indicating better performance.




The results are very interesting as we did not expect our AM2 systems to perform this well and to do so is very surprising when reviewing other media encoding results. In fact, we ran the test several times and verified our settings before accepting the test results. Perhaps the last few months of Core 2 Duo results constantly exceeding AM2's best by a wide margin have tainted our cognitive abilities.

Of course, we are using a mid-range AM2 processor against the budget C2D part (the AMD price cuts have helped matters there, as the price difference is currently only about $35) but we fully believe the majority of the performance difference lies in the chipset selection on the AM2 systems. It is only in the shrink and burn tests that we see the Intel platform flexing its muscles. The difference in the shrink portion of this test was minimal and mirrored our individual shrink test results. We noticed our Intel platform was able to feed data to our DVD drive in a consistent manner with each of the AM2 platforms suffering from stutters several times resulting in our drive slowing down during the burn process. It was difficult to figure out if this was a driver, disk controller, or CPU issue. We are still testing on other chipset and CPU combinations.

Our next test has us converting our day at the beach AVI file into a high definition WMV file suitable for our Aunt Gertrude to view on her new high definition LCD TV. We ensured our quality settings were set for High Definition output. The remaining options are left at default values and then we let this program do its magic. We are reporting the numbers in minutes/seconds to complete the conversion.


These results left our heads shaking and we busily scurried off to the far corners of the lab looking for answers. After several test repeats, a few new images, and a couple of drinks, we stand by the results. Our answer, which we came up with after we switched out the MSI G965 board for a Gigabyte G965 board and had the same results, is that our E6300 is just incapable of keeping up with our AM2 5200+ X2 in this test. Why? It's all about the cache and CPU speed. Our E6600 trounces the 5200+ X2 at stock speeds and scores near it when running at E6300 speeds thanks to the increased cache. Meanwhile, our E6400 is just slightly off the 5200+ X2 pace at a modest 300MHz increase in speed over the E6300.

Next on the list is our QuickTime Pro 7.1 test that will convert our newly downloaded .Mov file into a plasma screen pleasing H.264 format. We ensured our quality settings were set to their highest levels and then let the horses loose. The values reported are again in minutes/seconds for the conversion time, with lower numbers being better.


Once again, the AM2 platforms have a slight advantage with the 6150 scoring outside of the norm in this test - so far outside that we thought something was not kosher in the land of NVIDIA. Fortunately for us, the answer was easy once we reviewed the playback files. It seems as though the quality of our encoding results on the NVIDIA platform left much to be desired along with numerous frames missing during playback or other frames just tangled together at times. We ran this same test under XP without an issue so our first blip with Vista has now occurred. Our video images are presented below.

690G - Click to enlarge

G965 - Click to enlarge

6150 - Click to enlarge

In these particular screen shots both the G965 and 690G image quality results are basically even, but the 6150 is showing a case of the jaggies at full screen resolution. The entire playback file suffered from this issue.



Audio Encoding Performance

While the media encoding prowess of the Intel platform seemed to be lacking at times, we expect it to fair a little better on the audio side. Our audio test suite normally consists of Exact Audio Copy v095.b4 and LAME 3.98a3. We did run these tests with basically the same results but for our purposes today we will utilize iTunes 7.1 as it is one of the most utilized audio applications available due to the immense popularity of the iPod.

Our first task was to figure out what test CD to utilize. We needed one that contained a significant number of tracks and had over 600MB of data in order to properly stress our platforms. After rummaging around the lab and discounting the Burl Ives Greatest Christmas Hits album, we happened upon a CD that would work for our purposes as it was not working for anyone else, INXS Greatest Hits. This one time '80s glory masterpiece contains 16 tracks totaling 606MB of songs that was instantly whisked into our anxiously awaiting optical drive.

Our first test consists of utilizing iTunes to rip our INXS CD to WAV format on our hard drive. Even though CPU speed and optical drive selection have a great impact on these test results we are using the same optical drive between platforms for each test.


Our two AM2 platforms score slightly ahead of our Intel setup but the margin is minimal. This test was very consistent with each result never varying more than one second from the base score. Unlike the Nero Recode 2 DVD burning test, in this case our AMD systems did not suffer from drive issues.

Our next two tests have us utilizing iTunes to convert our WAV files into ACC or MP3 compatible formats. We utilize the 320kbps setting for both tests with the MP3 test also utilizing the variable bit rate option.



We usually see iTunes favoring the AM2 platforms in most test results. The results are no different here with both AM2 systems completing the ACC conversion about 50 seconds before the Intel system. In our MP3 conversion, the Intel system places second but the general differences are minimal. We attribute the 6150 performance to better storage controller and CPU throughput although standard memory bandwidth is lower than the other two boards.

Our final test has us creating a legal use copy CD from the base WAV files we collected when ripping the INXS Greatest Hits CD. Not that the world needs anymore of these CDs but it makes for a good closing test. We promise all CDs are destroyed after testing (much to the pleasure of our ears and those at the RIAA).


Once again our test results are extremely close with the optical drive playing a leading role in ensuring our results are fairly close. The minor differences are all attributable to the better disk performance of the NIVIDA 6150/430 combination in this case. We noticed each time that the disk access stopped about two to three seconds quicker with the 6150 platform. This is opposite of our Video Encoding results where each of our AM2 platforms could not keep the optical drive fed properly over the course of a burn session.



Media Performance

We will take a brief look at general media performance with our test suite for that includes Adobe Photoshop CS3, Adobe Photoshop Elements 5.0, and Windows Movie Maker.

We utilize the widely available Retouch Artists Photoshop Speed Test for measuring platform performance in Adobe's Photoshop CS3. The benchmark applies a number of actions to a test image while we measure the total time elapsed during the active portion of the test. We set history states to 1, cache levels to 4, and CS3 was configured to make use of all available system memory. The scores reported include the full conversion process and is represented in minutes/seconds, with lower numbers providing better performance.


The results are once again interesting as we did not expect our AM2 systems to perform this well. Photoshop has favored the Intel Core 2 Duo greatly in the past and once again we are looking at a speed issue with the E6300 that allows the AM2 systems to be in the hunt.

Our next test is one recommended by Intel, but the test itself appears to be fair and results are very repeatable. This test simply measures the amount of time required to fix and optimize 103 different photos weighing in at 63MB. Time is measured in minutes and seconds, with lower times resulting in better performance:


This test not only stresses the CPU but also gives the storage systems a good workout. Our 6150 platform barely edges out the Intel platform in a test where sheer CPU speed can make up the difference in elegant design. That sounds so familiar, are we talking about the differences between Athlon 64 and P4 NetBurst or Intel Core 2 Duo and Athlon 64 X2? Either way you cut it, the 6150 once again finishes ahead of the 690G due to better storage system performance.

Next on the list is our Windows Movie Maker test that will meld our European vacation content (recorded off TV) into a newly downloadable file that can be viewed on our Xbox 360 in a pleasing 720p format. The values reported are in minutes/seconds for the conversion time, with lower numbers being better.


Our Intel platform has a significant advantage in this test with the 6150 scoring better than the 690G in a test that stresses both the CPU and storage system again. These results were interesting as previous testing with this application indicated our AM2 platform would be a little more competitive.

File Compression Performance

In order to save space on our hard drives and ensure we had another CPU crunching utility, we will be reporting our file compression results with the latest version of WinRAR that fully supports multi-treaded operations and should be of particular interest for those users with dual core or multi-processor systems. Our series of file compression tests utilizes WinRAR 3.62 to compress our test folder that contains 444 files, ten subfolders, and 602MB worth of data. All default settings are utilized in WinRAR along with our hard drive being defragmented before each test.


Our G965 platform basically zips away from our two AM2 platforms in this test that heavily stresses the CPU subsystem and at times the storage system. The Core 2 Duo has always excelled in this benchmark and once again shows its strength. We were surprised that our two AM2 platforms scored this well with the advantage going to the NVIDIA 6150 due to better storage system throughput in our opinion.



Gaming Performance

As usual, gaming performance was tested with a variety of current games. We ran benchmarks at a 1024x768 resolution. Any more than this resulted in completely unplayable game conditions, although we did find several older non-FPS games such as the Civilization series, Railroad Tycoon 3, Rollercoaster Tycoon series, and others played acceptability at 1280x1024 with the quality settings at normal. We concentrated on Direct X games as OpenGL support under Vista is bad at this time.

Battlefield 2

This benchmark is performed using DICE's built-in demo playback functionality with additional capture capabilities designed in house. During the benchmark, the camera switches between players and vehicles in order to capture the most action possible. There is a significant amount of smoke, explosions, and vehicle usage as this a very GPU intensive Battlefield 2 benchmark. We run Battlefield 2 using medium quality graphics settings available in the video settings. The game itself is best experienced with average in-game frame rates of 35 and up.



Half-Life 2: Lost Coast

We use the built-in timedemo feature to benchmark the game. Our timedemo consists of starting at the bottom of the hill near the lake and ending in the old church. The Source engine timedemo feature is similar to the nettimedemo of Id's Doom 3 engine, in that it plays back more than just the graphics. The visual quality settings were set to high or medium where possible with HDR off. While the Source engine is notorious for giving great frame rates for almost any hardware setup, we find the game isn't as enjoyable if it isn't running at 35fps or above.



Company of Heroes

Company of Heroes was recently released and is proving to be a very addictive RTS game around the office. This game is very GPU intensive and also requires a hefty CPU at times. The game contains a built-in performance test that utilizes the game engine to generate several different action scenes along with a coffee argument as a sideline distraction before the war starts. We found the performance test gives a good indication of how well your system will perform throughout the game on average. Some of the in-game action sequences are more demanding than the performance test, but we generally found the game to be enjoyable with an average performance test frame rate above 30fps.



Sims 2

Sims 2 was released over two years ago and is constantly being updated with best selling expansion packs. In testing with the various expansion packs we did not notice any measurable differences in performance so our benchmark will be reflective of game play using the base game title. This particular game requires a decent CPU and very good GPU when utilizing the antialiasing, shadow, or high quality texture options.

We utilize FRAPS to capture the results from our replay file. Our benchmark consists of a three character scenario that takes approximately twenty minutes to set up and cycle through a series of daily events. The camera movements are varied as is the movement so we feel like this benchmark represents the typical game. We set most video options to high but disable shadows. We generally found the game to be enjoyable with an average frame rate above 30fps.



Gaming Summary

What else can you really say after reviewing the results? They are generally pathetic from a gaming perspective but the 690G obviously has some potential left in it at the lower resolutions. The 6150 performs okay considering the age of its core and we will see the new 6150SE and older 6100 chipset performing a few percent better overall but not enough to catch the 690G.

Considering the G965 was launched last fall and promised to bring about a difference in on-board video performance, we are still not seeing the results live up to the hype. The G965 was incapable of running Battlefield 2 at all and Half-Life 2 was an interesting slide show that allowed you to see every detail of the benchmark session in an excruciatingly slow way that would make a visit to the dentist for a root canal seem pleasant. However, G965 was able to generate decent results (for this grouping) in Company of Heroes and Sims 2, though the minimum frame rates in Sims 2 would bring the game to a slow crawl at times. Image quality generally favored the 690G in the majority of games we tried but the NVIDIA 6150 was close in most titles. The G965 had acceptable image quality but we could tell certain details were not as sharp or even evident when compared to the competing AM2 solutions.



Disk Controller Performance

The normal AnandTech iPeak test for measuring storage controller performance is designed to keep the hard drive as consistent as possible while varying the hard drive controller. The idea is to measure the performance of each hard drive controller with the same hard drive. Unfortunately, we discovered that iPeak does not work properly under Microsoft Vista at this time. We are currently looking at alternative test methods but will present HD Tune 2.53 results today.




The performance pattern according to these results would lead one to believe that the NVIDIA 6150/430 offers the worst disk performance as its minimum, maximum, and average transfer rates are the worst in the group, but it does have the best burst rates. The G965's ICH8 and 690G's SB600 Southbridge designs score identically in the first series of tests with the SB600 providing a higher burst rate at the expense of CPU utilization. In actual application testing, we generally found the NVIDIA 6150/430 offered the best overall disk performance.

Power Consumption

We measured power consumption at two states: at idle sitting at the Vista desktop and under load while running our 3DMark06 test. At both settings, Cool 'n Quiet/EIST were enabled to keep power consumption down to a minimum, although the biggest difference is made at idle.



In our idle tests we see that both AM2 platforms have an advantage over the Intel platform with or without power management turned on. In previous testing we know the Intel Core 2 Duo consumes less power than our 90nm 5200+ X2 samples, which indicates the G965 chipset is a power hog when compared to the AM2 solutions. The 690G scores extremely well in both our Idle and Load tests.



Image Quality

We will be providing additional image quality results in our mATX roundup that will provide in-game and Blu-ray screenshots, but for now here are a few samples that we collected during our recent testing.

Our first image quality test consists of playing back House of Flying Daggers utilizing Windows Media Player 11. We also used PowerDVD 7 and WinDVD 8 with basically the same results.

690G - Click to enlarge

G965 - Click to enlarge

6150 - Click to enlarge

We think the 690G and G965 provide the best overall picture with the 6150 just not producing an image with the same color and detail as the other platforms. Utilizing the PureVideo application resulted in a slightly better picture but still behind the 690G and G965. Overall, we tend to think the G965 had the best picture quality during playback but your mileage will vary.

Our second image quality test consists of playing back a Windows WMV-HD clip from the Fighter Pilots DVD utilizing Windows Media Player 11. We also tested with WinDVD 8 and it generated similar results.

690G - Click to enlarge

G965 - Click to enlarge

6150 - Click to enlarge

We believe the 690G and G965 once again provide the best overall picture quality; the 6150 produces an image that is clear but without the color saturation of the other two alternatives. These results could be considered a toss up depending on personal preferences. We tend to think the G965 has the best picture quality although it took a lot of long stares for our opinion to be formed on the subject. We think the pilot's harness was slightly sharper in the G965 picture thus swaying our opinion.

Our final image quality test consists of playing back a Windows .AVI clip from our encoding tests utilizing Windows Media Player 11. We also tested with WinDVD 8 and PowerDVD 7 with both generating similar results. The results from the encoding process from the AVI to WMV file format will be provided shortly. It is near impossible to capture the image at the exact same point with Windows Media player but the images in this test group are very close to being identical.

690G - Click to enlarge

G965 - Click to enlarge

6150 - Click to enlarge

When reviewing these images we firmly believe the G965 has the best overall image from both a color saturation and picture detail viewpoint. The 690G and 6150 produce an almost equal image that but we think the clarity in the 6150 images give it an edge.

In our video playback tests we believe the G965 had the best picture quality of our solutions with the 690G performing equally in most cases and the 6150 trailing behind. Intel touted the G965's Clear Video Quality at product launch and from all indications this technology has made a difference, at least in our limited testing to date.



Final Words

AMD has produced a very good integrated graphics platform with the 690G chipset, there is no doubt that this platform is the first step from the company in their goal to fuse core logic, graphics, and processing into a neat and tidy single package offering. We see this release as the first shot across the bow of the SS Intel and HMS NVIDIA that signals AMD's intent to play seriously in the OEM and business sectors. Considering the IGP market accounts for almost 90% of PC shipments, then it becomes obvious where AMD wishes to concentrate its efforts at this time.

The 690G features very low power consumption, class leading X1250 graphics performance with the Avivo video processing engine, HDMI output with full HDCP support, excellent audio and video output capabilities, and competitive performance all wrapped in a $80 price tag. It seems to us the 690G is destined to find a home in dedicated home theater systems or for those owners needing a mainstream system centered on multimedia capabilities.

While we are impressed with this release there are also some concerns. The performance of the X1250 graphics core was certainly class leading but one look at the competition tells an ugly truth for certain users. Despite being faster than the NVIDIA GeForce 6150 and Intel G965 in games, the system overall is still incapable of providing a decent gaming experience for even the casual gamer. It's still very underpowered for recent titles unless you consider 800x600 gaming to be a pleasurable experience.

If you're willing to turn the majority of detail levels down to their minimum, then yes, you can play several popular games such as Sims2 at 1280x1024. This resolution is the standard base level now as a significant amount of monitors sold over the past few years use 1280x1024 as their native resolution. The fact that we have an integrated graphics platform being released in 2007 that can barely play most top titles delivered in 2005 at 1024x768, much less 1280x1024, is an issue with us. We understand these platforms are mainly directed to the email, home/office, and Internet crowds, but those wanting to dabble with the occasional game will end up having a sour experience unless Internet Poker or Bejeweled is your game of choice.

That is our main issue, the 690G, G965, and 6150 are targeting the all-in-one home user market and not the office sector. As such we expect a platform that is truly capable of running most game titles decently while performing other duties such as media/audio playback, digital content creation, email, and handling your tax returns with aplomb. When the marketing information tells us we can expect our system to do this, then that is point where our expectations are set. The office sector relies upon the 690V, Q965, and 6150SE chipsets and low-end processor choices to provide just enough functionality to keep the typical office user productive.

With the latest round of processor price cuts and an $80 price tag for the motherboard, we find AMD has a very competitive platform in comparison to the Intel Core 2 processor and G965 platform. Unfortunately, this means that you must utilize a mid-range AM2 processor in order to get performance comparable to that of Intel's base Core 2 processor at similar prices. Our AMD motherboard and processor will cost around $314 compared to $298 for the G965 and Core 2 Duo E6300. If the multimedia capabilities of the 690G are really important then paying the extra $16 for almost equal performance in a wide range of applications should be worth your consideration.

A couple of intangibles that we found during testing swayed us to the AM2 platform to some degree. The overall maturity of the platform, ease of setup, native IDE ports, and platform stability during testing impressed us. While the sheer power of the Intel Core 2 Duo still holds us captive, we found the current AM2 platforms offer a compelling alternative in the low end sector where price to performance comparisons mean everything. This could all change depending upon Intel's pricing structure and new IGP chipset releases over the coming months but for now, we would seriously consider an AM2 system for general desktop or multimedia usage.

Overall, the 690G is a very capable chipset, but one that is late to market in our opinion. We were impressed by its multimedia prowess and features; certainly enough to whisk it into our HTPC test bed, but are concerned about its timing. AMD has not taken that next real step in performance for an integrated graphics platform in terms of performance. This release just barely distances itself from the NVIDIA 6150 series that has been on the market for almost 18 months. In terms of features, NVIDA will soon be releasing the GeForce 7050 that fully supports PureVideo, HDMI with HDCP support, and additional core tweaks that should result in performance closer to that of the 690G. Intel is busily working on their new G35 chipset with native HDMI and hopefully improved game performance although we doubt it unless their drivers improve significantly. While AMD can claim the current IGP performance crown, they need to have another solution ready quickly before NVIDIA or Intel comes calling for it.