Overclocking

Asus A8R-MVP Overclocking Testbed
Processor: Athlon 64 4000+
(2.4GHz, 1MB Cache)
CPU Voltage: 1.45V (default 1.40V)
Cooling: Thermaltake Silent Boost K8 Heatsink/Fan
Power Supply: OCZ Power Stream 520W
Memory: OCZ PC3200* Platinum Rev. 2
(Samsung TCCD Memory Chips)
*The current equivalent OCZ memory is OCZ PC4800
Hard Drive: Seagate 120GB 7200RPM SATA 8MB Cache
Maximum OC:
(Standard Ratio)
246x12 (4x HT, 2.5-3-3-7)
2952MHz (+23%)
Maximum FSB:
(Lower Ratio)
325 x 9 (3x HT, 2T*, 3-4-3-8)
(2925MHz, 2 DIMMs in DC mode)
(+62.5% Bus Overclock)

Screen captures on the first page have already announced that the Asus A8R-MVP has set new overclocking records in our testing. We actually found 330 to be stable in memtest86 and booted into Windows at a little more than 330 with memory set to a DDR400 starting point. However, Windows was not completely stable at 330 and required a slight lowering to 325 (DDR650). The 325 setting was completely stable, ran our benchmarks, and ran for more than 2 days on just air cooling.

Our first efforts at overclocking the A8R-MVP ran into a road block at just over 260. We have found some Asus boards in the past that did not like overclocks to be immediately set to high values, so we started again at 250. By going up just 5 to 10FSB at a time, we were able to reach 325. Since we had no idea that we would reach such a high OC on this mainstream board, we have screen captures at every 5MHz from about 270 to 330. Our advice to overclockers on this board is to move the overclock up slowly.

*UPDATE: After further testing we have concluded that the A8R-MVP handles 1T timings only to about 260-265. Above 260-265 you will need 2T timings to reach 325 or so clock frequency. We ran into a wall at 260 and apparently the BIOS reset. The default is 2T and we didn't catch the Command Rate reset. This board uses very aggressive timings, and the 2T memory bandwidth was so fast we did not realize the board was actually running at 2T. This does not change the fact that 325 at 2T is an outstanding overclock, but it is achieved at 2T timings, and not at 1T as originally reported.

Those who don't understand overclocking or who don't want to bother will find extensive automatic overclocking options in the Asus BIOS. This allows you to set an overclock and have the board adjust the related settings. These work well for modest overclocking, but they will not allow the extreme results achieved manually on the Asus A8R-MVP.

Memory Stress Testing

The Asus A8R-MVP easily handles 2-2-2-7-1T timings at stock speed, as do almost any of the current boards for AMD Socket 939 from NVIDIA, SiS, VIA, ATI, and ULi. However, the default setting for Command Rate in the Asus BIOS is 2T, and you need to manually set Command Rate to 1T for best performance with 2 DIMMs in Dual Channel mode. Asus tells us that this was done to accommodate some memory manufacturers that had asked for a default 2T Command Rate for maximum compatibility with their memory. Once set, we had no issues using 1T command Rate with any memory that we tested.

Running four double-sided 512MB or 1GB DIMMs is much more demanding than running two DS DIMMs, and Asus did not have any extra magic here. Like every board that we have tested except the DFI RDX200, we needed to drop the Command Rate to 2T with 4 DS DIMMs. With 4 DIMMs, the A8R-MVP remained stable with the same aggressive 2-2-2-7 timings used for two DS DIMMs.

Stable DDR400 Timings - 4 DIMMs
(4/4 DIMMs populated)
Clock Speed: 200MHz
CAS Latency: 2.0
RAS to CAS Delay: 2T
RAS Precharge: 7T*
Precharge Delay: 2T
Command Rate: 2T
*7T was determined by MemTest86 benchmarks to deliver the widest bandwidth with the ATI Rx480/482 chipset. While the board would operate at tRAS of 5T or lower, all benchmarks were run at 7T.
Basic Features: Asus A8R-MVP Test Setup
Comments Locked

74 Comments

View All Comments

  • poohbear - Thursday, November 24, 2005 - link

    thanks for clarifying.:) guess it's best to wait and see other reveiws to get a general consensus though.
  • Calin - Wednesday, November 23, 2005 - link

    Maybe the design team had a set of design cues for the board, received from the top (management) - these include cost, performance, reliability, and possibly even other metrics. The ATI chipset might have been too good for those cues, or maybe they worked to optimize the board - and told nothing to management about their substantial successes.
    Or maybe the board was a lucky one, and maybe 1 in 100 will get near those results, and the rest will be mainstream. Anyway, I think it is a good design, and not a lucky board

    Calin
  • poohbear - Wednesday, November 23, 2005 - link

    jesus 2900mhz on air?!!?? i LOVE competition.:)
  • CrystalBay - Wednesday, November 23, 2005 - link

    Hi Wes, Did you run into any coldboot problems ?
  • Wesley Fink - Wednesday, November 23, 2005 - link

    No Coldboot problems at working overclocks. We did find, however, that failed overclocks at very high frequencies usually required the system be turned off (power switched off) for the board to recover from the failed OC. This is common to many motherboards, but it is always easier if the board recovers on a warm boot.
  • Diasper - Wednesday, November 23, 2005 - link

    Just wondering why there weren't any game benchmark comparisons against the Asus A8N32 given the article set out by comparing them and given the numerous benchmarks with it including Aquamark, how come it was excluded from the rest of the game benchmarks?? It would be pretty nice/important if you included them.
  • nvidia4ever - Wednesday, November 23, 2005 - link

    I was wondering the same. Either the results are not favorable for ATI or there are results coming that were not posted yet. Either way it looks like a great board.
  • Wesley Fink - Wednesday, November 23, 2005 - link

    The original benchmarks with the A8N32-SLI Deluxe were run at 1600x1200 with AA/AF on. This is because we were trying to determine if Dual x16 SLI made any difference compared to Dual x8 SLI. When we tested the ATI we reran benchmarks on one of the current top nForce4 boards - the DFI LANParty nForce4 SLI - to provide a comparison with the nForce4 SLI chipset. We did not have the Asus A8N32-SLI available for retesting since another reviewer was using the board for benchmarking. It also didn't seem important to also rerun benches on the A8N32-SLI since its performance was comparable to the DFI nF4 SLI.

    At any rate, the A8N32-SLI is available again and we will rerun some benches at 1280x1024 and add results to the game graphs. For "Standard Score" benchmarks like 3DMarks and Aquamark 3 the video resolutions are always the same and the A8N32-SLI results are already included in the graphs.
  • Beenthere - Wednesday, November 23, 2005 - link

    It was pretty obvious a year ago that ATI had created a very impressive chipset for the AMD Mobos. The Southbridge ULi chip resolves perceived deficiencies by those who believe they need faster USB and S-ATA 2, even though in reality this simply isn't true. The fact that the ATI Northbridge chipset O/C's so well is proof of an excellent design and one to build a Helleva reputation on in the Mobo market. The ATI chipset makes it pretty obvious that you don't need to endure high prices and marketing gimmicks of some Mobo mfgs. to promote over-priced "gamers or overclockers" Mobos that don't deliver as much, let alone more PC performance than the ATI chipset Mobos, which include all the practical overclocking BIOS options anyone would possibly need.
  • sunshine - Wednesday, November 23, 2005 - link

    RE: "Our first efforts at overclocking the A8R-MVP ran into a road block at just over 260. We have found some Asus boards in the past that did not like overclocks to be immediately set to high values, so we started again at 250. By going up just 5 to 10FSB at a time, we were able to reach 325."

    What is it exactly that prevents you from setting up the overclocking settings all at once? Why must you increase at only 5 - 10 mhz at a time? Is the Asus Bios boobytrapped to prevent someone from frying their motherboard or CPU???

Log in

Don't have an account? Sign up now