facebook rss twitter

Review: Shuttle AK37 GT/R KT400 Motherboard

by Tarinder Sandhu on 19 October 2002, 00:00

Tags: Shuttle, AMD (NYSE:AMD), VIA Technologies (TPE:2388)

Quick Link: HEXUS.net/qanh

Add to My Vault: x

Benchmarks I

Benchmarking, as always, starts off with SiSoft SANDRA's memory benchmark.

I'll take a minute to explain these numbers, as they will give us an insight into later benchmarks. The Pentium 4, with its quad-pumped FSB, is able to use the 2.7GB/s on offer to it. In fact, at 133FSB, the P4 can use up to 4.26GB/s of bandwidth.

The Athlon XP, on the other hand, with its double-pumped FSB, can only use 2.1GB/s of memory bandwidth at 133FSB. So, running faster memory will not give greater scores, we're saturating the Athlon's capacity to use memory with just 2.1GB/s.

Why are the DDR400 scores lower than DDR333 ?. The answer lies in the fact that we don't gain much by running anything other than DDR266 memory at 133FSB. To run at DDR400, I have to lower the timings which were used when benchmarking at DDR333. DDR333 memory does have a slightly advantage over DDR266 memory in that we have a little bandwidth left over for system resources once we supply 2.1GB/s to the processor. Running DDR400 memory on a processor stuck on a double-pumped 133FSB is pointless, seemingly.

Now to the unbuffered benchmark. Here no prefetching is allowed to take place.

Once again, having to use weaker timings at DDR400 manifest themselves in lower scores. The XP simply cannot use the potential of DDR400 memory that is on offer, and what it does use is compromised by slower timings. That's why strict timings at DDR333 give a higher benchmark. The P4 once again leads the way.

Let's see how these memory differences effect our first benchmark, Pifast. It simply calculates the constant Pi to the desired number of decimal places. I've chosen 10 million using the fastest method possible. Memory bandwidth (or how much you can use) and timings play an important part in this benchmark.

Here, again, the inability of the XP to use much more than 2.1GB/s at 133FSB, coupled with the lower timings that have to be used when running at DDR400, show themselves with a slower time than DDR333 with strict timings. The Athlon XP's fantastic FPU ousts the P4 2.26GHz's performance.