Abit AW9D-MAX: When "Beta" MAX is a good thing
by Gary Key on September 8, 2006 3:10 AM EST- Posted in
- Motherboards
Memory Performance
We have been saying for years that the Buffered benchmark does not correlate well with real performance in applications on the same computer. For that reason, our memory bandwidth tests have always included an Unbuffered Sandra memory score. The Unbuffered result turns off the buffering schemes, and we have found the results correlate well with real-world performance as we will see shortly.
With the E6600, our Abit board offers a 5% improvement in the stock clock speed Sandra Unbuffered test and a 3% improvement in the overclock tests over the ASUS board. The ASUS board holds a small advantage in both clock settings in our latency tests which is surprising considering the Abit advantage in the Unbuffered tests.
General Performance
We also tested a couple of real world applications that typically stress the CPU, memory, and storage systems along with a synthetic test to see if the performance differences in our memory synthetic tests carry over to the desktop. Our real world application tests include activities that are common on the desktop.
Our first test was to measure the time it takes to shrink the entire Office Space DVD that was extracted with AnyDVD into a single 4.5GB DVD image utilizing Nero Recode 2. Our second test consists of utilizing Exact Audio Copy as the front end for our version 3.98a3 of LAME. We set up EAC for variable bit rate encoding, burst mode for extraction, use external program for compression, and to start the external compressor upon extraction (EAC will read the next track while LAME is working on the previous track, thus removing a potential bottleneck with the optical drive). Our test CD is INXS Greatest Hits, a one time '80s glory masterpiece containing 16 tracks totaling 606MB of songs. The results of our tests are presented in minutes/seconds with lower numbers being better.
Our third test is Cinebench 9.5 which heavily stresses the CPU subsystem while performing graphics modeling and rendering. We utilize the standard benchmark demo within the program along with the default settings. 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. The results are presented in a standardized score format with higher numbers being better.
Our fourth test is 3DMark06 which tests the graphics and CPU subsystems. The 3DMark series of benchmarks by Futuremark are among the most widely used tools for benchmark reporting and comparisons. Although the benchmarks are very useful for providing apples-to-apples 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. The results are presented in a standardized score format with higher numbers being better.
The performance of the Abit AW9D-MAX was very consistent and in alignment with the memory test results. We found the board to be very responsive and extremely stable during testing. In fact, if you were doing a blind box test it would have been difficult to figure out which board was performing the best. The differences in performance between the two boards are very minor but the Abit board had up to a 5% advantage in our video/audio encoding tests. This indicates to us Abit has properly optimized the processor, memory, and storage subsystems within their BIOS code. However, this is a beta BIOS and performance could change either way. We suggest waiting on the production level BIOS before drawing any absolute performance conclusions. Let's see if these results carryover into our game benchmarks.
 |
| Click to enlarge |
We have been saying for years that the Buffered benchmark does not correlate well with real performance in applications on the same computer. For that reason, our memory bandwidth tests have always included an Unbuffered Sandra memory score. The Unbuffered result turns off the buffering schemes, and we have found the results correlate well with real-world performance as we will see shortly.
With the E6600, our Abit board offers a 5% improvement in the stock clock speed Sandra Unbuffered test and a 3% improvement in the overclock tests over the ASUS board. The ASUS board holds a small advantage in both clock settings in our latency tests which is surprising considering the Abit advantage in the Unbuffered tests.
General Performance
We also tested a couple of real world applications that typically stress the CPU, memory, and storage systems along with a synthetic test to see if the performance differences in our memory synthetic tests carry over to the desktop. Our real world application tests include activities that are common on the desktop.
Our first test was to measure the time it takes to shrink the entire Office Space DVD that was extracted with AnyDVD into a single 4.5GB DVD image utilizing Nero Recode 2. Our second test consists of utilizing Exact Audio Copy as the front end for our version 3.98a3 of LAME. We set up EAC for variable bit rate encoding, burst mode for extraction, use external program for compression, and to start the external compressor upon extraction (EAC will read the next track while LAME is working on the previous track, thus removing a potential bottleneck with the optical drive). Our test CD is INXS Greatest Hits, a one time '80s glory masterpiece containing 16 tracks totaling 606MB of songs. The results of our tests are presented in minutes/seconds with lower numbers being better.
Our third test is Cinebench 9.5 which heavily stresses the CPU subsystem while performing graphics modeling and rendering. We utilize the standard benchmark demo within the program along with the default settings. 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. The results are presented in a standardized score format with higher numbers being better.
Our fourth test is 3DMark06 which tests the graphics and CPU subsystems. The 3DMark series of benchmarks by Futuremark are among the most widely used tools for benchmark reporting and comparisons. Although the benchmarks are very useful for providing apples-to-apples 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. The results are presented in a standardized score format with higher numbers being better.
 |
| Click to enlarge |
The performance of the Abit AW9D-MAX was very consistent and in alignment with the memory test results. We found the board to be very responsive and extremely stable during testing. In fact, if you were doing a blind box test it would have been difficult to figure out which board was performing the best. The differences in performance between the two boards are very minor but the Abit board had up to a 5% advantage in our video/audio encoding tests. This indicates to us Abit has properly optimized the processor, memory, and storage subsystems within their BIOS code. However, this is a beta BIOS and performance could change either way. We suggest waiting on the production level BIOS before drawing any absolute performance conclusions. Let's see if these results carryover into our game benchmarks.
Facebook
Twitter
RSS
56 Comments
View All Comments
Jedi2155 - Friday, September 8, 2006 - link
Oh heck yea! I thought I was going to have to go with Asus P5B deluxe board mainly due to the color scheme as I'm a bigger fan of Black & Blue than Black & Red. Too bad for all those other people who prefer red tho.wolf550e - Friday, September 8, 2006 - link
How much is Scyhte paying you?Madellga - Friday, September 8, 2006 - link
Actually his review is very neutral and not a PR stunt.This space is to discuss the review itself and the product there.
If you want to make such comments I suggest paging him, instead of writing here on the open.
Gary Key - Friday, September 8, 2006 - link
Actually I paid NewEgg $51.99 for the pleasure of using the Scythe Infinity. :) Scythe does not advertise with us and the picture was published based upon numerous reader requests wanting to know how the larger heatsinks fit on the boards. I still love and use the Tuniq 120 but until they are readily available again my current air cooling favorite is the Infinity. Just in case it comes up, the E6600 was also bought from NewEgg and is not supplied by Intel.
DigitalFreak - Friday, September 8, 2006 - link
LOL! First time I've ever heard a motherboard described as "menacing".
mostlyprudent - Friday, September 8, 2006 - link
I am very pleased to see Abit producing a strong board again. Two of my older systems are still running with Abit boards (4 years old and 2.5 years old) with no issues. I am still deciding, but the PCI slot issue is a tough pill to swallow.BTW, there is a type-o in the last paragraph on page 3 "Although this 'typcially' worked..."
GoatMonkey - Friday, September 8, 2006 - link
I used to be a big Abit fan. I bought 5 Abit motherboard for myself over the years, and built at least 4 other systems for friends with Abit motherboards. Unfortunately, over half of them failed after several months of use. Abit really needs a good warranty and some good testimonial of quality to get me back.yyrkoon - Friday, September 8, 2006 - link
ABIT has a decent warranty policy, the only problem is that they exchange 're-certified' boards for your new one.We've had to deal with ABITs RMA a couple of times in the last two years, and while they did replace the boards, the process was slow, and again, they replace it with a re-certified board. However, it turned out it wasnt the motherboard that was bad at all, but a ATI videocard (pre-PCIE, and additional card power), that was drawing too much power from the AGP slot.
ABIT forums, while not owned or paricipated by any ABIT workers (that I know of), is second to none. If you cannot find someone on ABITs forums to help with an issue, then said issue is rare, or hard to trace.
I'm finding that more, and more, that motherboards dont really go bad (short term), but often 'broken' motherboards are configured improperly by the user, that has limited experience with that brand, or a user that really hasnt a clue how to properly setup a motherboard. This doesnt include the rare chance of a dead out of the box motherboard, or the random hard to troubleshoot other than motherboard issues, and I've recently experience the latter here myself (an Asrock board that would lock up within three days, no BSoD, and nothing standard fixed the problem).
The main reason why I like ABIT, is that usually ABIT boards have stability that is second to none, and they perform very well.
granulated - Friday, September 8, 2006 - link
I know that the Scyhte Infinity is approx 12cm x 12cm but blimey !It's looks massive !
yyrkoon - Friday, September 8, 2006 - link
Gary, was wondering if yo ucould confirm if ABIT boards with eSATA, and a SIL 3132 controller will in fact work with a SATA port multiplier. From all the researching Ive done for the last year or so would indicate so, but I would like ot make sure before investing loads of cash in an external RAID 5 array, only to have it not work.