What to watch for in the test results
As with all chipset reviews, we're testing to see which is fastest, but that doesn't mean you should expect wide variances in performance. We're not dealing with fundamentally different technologies, for the most part, so they ought to run pretty close to one another. Also, remember that very small differences don't really matter, because there's some margin of error to our tests. Yes, we do run the tests multiple times and average the scores, but still, many differences are statistically insignificant. And in real-world use, even larger margins of victory are probably going to be imperceptible. Keep that in mind.

What you'll want to look out for is whether the KT333 and SiS 745 can take some advantage of DDR333 memory. Since the KT333 will be a drop-in replacement for the KT266A, it doesn't really have to run circles around the 266A to justify its existence. Ditto for the 735/745.


VIA's KT333 reference motherboard

Our testing methods
As ever, we did our best to deliver clean benchmark numbers. Tests were run at least twice, and the results were averaged.

Our test systems were configured like so:

AMD 760 NVIDIA
nForce 420D		 SiS 735 SiS 745 VIA KT266 VIA KT266A VIA KT333
Processor AMD Athlon XP 2000+ 1.67GHz
Front-side bus 133MHz (266MHz double-pumped) 133MHz (266MHz double-pumped) 133MHz (266MHz double-pumped) 133MHz (266MHz double-pumped) 133MHz (266MHz double-pumped) 133MHz (266MHz double-pumped) 133MHz (266MHz double-pumped)
Motherboard Gigabyte GA-7DX MSI K7N420 Pro ECS K7S5A ECS K7S6A Shuttle AK31 rev. 2 VIA KT266A reference VIA KT333 reference
North bridge AMD 761 nForce 420D 735 745 VT8366 VT8366A VT8367
South bridge VIA VT82C686B nForce MCP VT8233 VT8233 VT8233A
Memory size 256MB
(1 DIMM)		 256MB
(2 DIMMs)		 256MB
(1 DIMM)		 256MB
(1 DIMM)		 256MB
(1 DIMM)		 256MB
(1 DIMM)		 256MB
(1 DIMM)
Memory type Micron PC2100 DDR SDRAM Micron PC2100 DDR SDRAM Micron PC2100 DDR SDRAM Micron PC2100 DDR SDRAM
KingMax PC2700 DDR SDRAM		 Micron PC2100 DDR SDRAM Micron PC2100 DDR SDRAM Micron PC2100 DDR SDRAM
KingMax PC2700 DDR SDRAM
Graphics NVIDIA GeForce3 Ti 500 64MB (Detonator XP 21.83 video drivers)
Sound Creative SoundBlaster Live!
Storage Maxtor DiamondMax Plus D740X 40GB 7200RPM ATA/133 hard drive
(Drives courtesy of KickAss Gear)
OS Microsoft Windows XP Professional
OS updates None

We chose to test with VIA's K266A reference motherboard because it's the same board as the KT333 reference mobo. Using the reference boards will allow a true chipset-to-chipset comparison. We know from past experience that VIA's KT266A reference board performs very much like production KT266A boards, as well.

To keep things fair, we tested with the memory timings set as aggressively as possible on each motherboard, so long as the system was stable. Because we were using very high quality memory, using aggressive memory timings generally wasn't a problem. The one exception was the SiS 745, which simply refused to POST if we set the CAS latency to 2, regardless of whether the memory was running at 266MHz or 333MHz. We tried it with several different DDR333 DIMMs, and the ECS 745 board wouldn't POST at CAS 2 with any of them, even with the memory clock set to DDR266. So in the case of the 745 chipset, we had to test at CAS 2.5.

Also, note that we tested the nForce with two 128MB DIMMs in order to make sure both of its memory banks were populated.

The test systems' Windows desktops were set at 1024x768 in 32-bit color at an 85Hz screen refresh rate. Vertical refresh sync (vsync) was disabled for all tests.

We used the following versions of our test applications:

All the tests and methods we employed are publicly available and reproducible. If you have questions about our methods, hit our forums to talk with us about them.