Gigabyte 7VAXP (KT400): Breathing Some Life Back into KT400
by Evan Lieb on October 3, 2002 12:30 PM EST- Posted in
- Motherboards
Stress Testing the Gigabyte 7VAXP
Like we have done with the ASUS A7V8X, ECS L7VTA, and Soyo KT400 Ultra, the Gigabyte 7VAXP had to go through some massive DDR333 and DDR400 testing before we could truly dub this the "fastest" Socket A motherboard on the market. Luckily, Gigabyte did an exceptional job tweaking the 7VAXP for DDR333 AND DDR400 performance versus the other KT400 motherboards we've seen on the market.
Here are the memory timings we were able to run with DDR333 filled in all memory
banks running Prime95 for 48 hours straight. We also ran our benchmark suite
for several hours to make absolutely sure the 7VAXP was stable at these settings:
Stable
DDR333 Timings |
|
Clock
Speed:
|
166MHz
|
Timing
Mode:
|
Turbo
|
CAS
Latency:
|
2
|
Bank
Interleave:
|
4-bank
|
Precharge
to Active:
|
2T
|
Active
to Precharge:
|
5T
|
Active
to CMD:
|
2T
|
Command
Rate:
|
1T
|
The Gigabyte 7VAXP tied with the Soyo KT400 Ultra for the best DDR333 timings
we've ever been able to achieve using Corsair CL2 DDR400 memory. There's an
option for Ultra DRAM timing mode, but that was not stable enough for our liking.
Our DDR333 performance results for the Gigabyte 7VAXP and Soyo KT400 Ultra motherboard
lead us to believe that KT400 boards are marginally faster than VIA's previous
Socket A chipset, the KT333.
DDR400 was equally impressive for the most part. With all banks filled, we only
had trouble running at the most aggressive timings possible (i.e. similar to
DDR333 settings). Once we backed off the timings however, we were able to run
DDR400 with all banks filled operating at the following settings:
Stable
DDR400 Timings |
|
Clock
Speed:
|
200MHz
|
Timing
Mode:
|
Normal
|
CAS
Latency:
|
2.5
|
Bank
Interleave:
|
2-bank
|
Precharge
to Active:
|
3T
|
Active
to Precharge:
|
5T
|
Active
to CMD:
|
3T
|
Command
Rate:
|
1T
|
These settings aren't stellar, but in all reality they aren't that bad either, as the fastest timings compared to the ones above won't increase performance noticeably anyway. You will see a performance drop due to the lowered bank interleave and Precharge to Active/Active to CMD settings, but it won't be more than 10%, which is the point at which a performance delta is noticeable in real world applications. We should also note that this is the only KT400 motherboard that can run DDR400 with all banks filled, and so therefore we were quite satisfied with our results, despite the tiny performance boost the DDR400 afforded.
0 Comments
View All Comments