The Return of 2-2-2: Corsair 3200XL & Samsung PC4000
by Wesley Fink on June 15, 2004 9:00 PM EST- Posted in
- Memory
AMD Athlon 64 Performance Test Configuration
The recently introduced nForce3-250 and VIA K8T800 PRO chipsets finally promise a working PCI/AGP lock to Athlon 64. Earlier chipsets only had a working AGP/PCI lock on a handful of boards such as the AOpen AK89 Max. Thus far, we have tested the Epox, MSI, Chaintech, and Gigabyte nF3-250 chipset motherboards and found a working AGP lock on all these Socket 754 boards. We also found a working PCI lock on the 2nd revision of the Abit KV8 PRO based on the VIA K8T800 PRO chipset, but we had some issues with multipliers on that board.The same chipsets are used with the just-released AMD Socket 939 Dual-Channel processor. We are in the early stages of testing dual-channel Socket 939 motherboards, but we have been working with Asus on their A8V Deluxe based on the VIA K8T800 PRO chipset. Early revisions had no PCI/AGP lock and limited overclocking, but Asus made a hardware revision to shipping boards, which added a working AGP/PCI lock. The board has also improved through a number of beta BIOS to the point where the most recent beta BIOS has fixed many of the issues with overclocking on the Asus A8V. We have been able to achieve 1:1 overclocks as high as 265 FSB on the A8V with the latest BIOS.
While it is far too early to establish a standard memory test bed with a Socket 939 board, we have been experimenting with the working AGP/PCI lock to allow effective testing of Athlon 64 memory performance. With Intel moving to DDR2 in the upcoming 915/925X, we will likely move DDR testing to an Athlon 64 Dual-Channel test bed in the near future.
| Athlon 64 Performance Test Configuration | |
| Processor(s): | AMD Athlon 64 3800+ Socket 939 |
| RAM: | 2 x 512MB Corsair 3200XL (DS) 2 x 256MB Samsung PC4000 (SS) |
| Hard Drives | Seagate 120GB IDE 7200 RPM (8MB Buffer) |
| PCI/AGP Speed | Fixed at 33/66 |
| Chipset Drivers: | VIA Hyperion 4.51 |
| Video Card(s): | ATI 9800 PRO 128MB, 128MB aperture, 1024x768x32 |
| Video Drivers: | ATI Catalyst 4.6 |
| Power Supply: | Antec True Power 430W |
| Operating System(s): | Windows XP Professional SP1 |
| Motherboards: | Asus A8V Deluxe (VIA K8T800 PRO) Beta BIOS 1005.020 |
In our testing with Socket 939 boards, we found the best performance is achieved at a tRas setting of 10. All Performance Tests were run with a 10 tRas setting.
Test Settings
The Athlon 64 also has the unique feature of unlocked multipliers below the rated speed on all processors, and both above and below rated speed on the FX chips. This feature is not currently available on Pentium 4 processors. The unlocked lower ratios combined with a working AGP lock makes it possible to take a different approach to testing memory performance on the Athlon 64. It is possible to look at a fixed Processor speed and variable Memory speeds to see the real impact of just higher memory speeds on typical performance.The standard Quake 3 (Open GL), Super PI (raw calculation performance), SiSoft Sandra 2004 Standard Buffered memory test (synthetic memory test), and SiSoft Sandra 2004 Standard UNBuffered memory test were run as usual. However, to test more effectively the effect of memory speed on performance, we expanded the benchmarks used for testing. UT2003 (Direct X 8) and Aquamark 3 (Direct X 9) were added to the memory tests to provide a broader range of performance measurements.
Facebook
Twitter
RSS
11 Comments
View All Comments
Pumpkinierre - Tuesday, June 15, 2004 - link
Good article again, Wesley. Pity its not DDR500 at 2-2-2. I'm still holding off upgrading. I wouldnt trust that VIA chipset with the Corsair RAM. Plenty of people run their memory outside SPD specs withot problems. And don't give up on the i875 yet. There's a lot of issues with DDR2 and 915/925. I notice that ABIT have brought out a Sckt775 865 mobo. Hmm I wonder why?Also the P4/i875 seems to equal or better the S939 a64 in unbuffered sandra which I wouldnt have expected because of the on die a64 mem. controller etc.. Then in the buffered test the a64 clearly gets the upper hand which again is a suprise as many of the buffers are associated with MMX/SSE/SSE2 where the a64s are supposed to be weaker. I only trust the unbufferd tests but this may explain the fact that the FX chips beat the P4s on memory bandwidth but were behind on the bandwidth intensive encoding tests.
In the one test (Samsung mem.) where you test the a64 at different bus speeds (200&240MHz), the gaming results were equal or worse in the game tests despite an ~85 increase in mem. bandwidth ! Unfortunately you had different memory timings but it reinforces the importance of latency reduction rather than bandwidth for gaming performance.