8 - Queen CPU test | Memory performance

Queen CPU test

This simple integer benchmark focuses on the branch prediction capabilities and the misprediction penalties of the CPU. It finds the solutions for the classic "Queens problem" on a 10 by 10 sized chessboard. At the same clock speed theoretically the processor with the shorter pipeline and smaller misprediction penalties will attain higher benchmark scores. For example - with HyperThreading disabled - the Intel Northwood core processors get higher scores than the Intel Prescott core based ones due to the 20-step vs 31-step long pipeline. However, with HyperThreading enabled the picture is controversial, because due to architectural bottlenecks the Northwood core runs out of internal resources and slows down. Similarly, at the same clock speed AMD K8 class processors will be faster than AMD K7 ones due to the improved branch prediction capabilities of the K8 architecture.

CPU Queen test uses only the basic x86 instructions, it consumes less than 1 MB system memory and it is HyperThreading, multi-processor (SMP) and multi-core aware and thus is a multi-threading CPU Benchmark with MMX, SSE2 and SSSE3 optimizations.

Traditionally a benchmark where AMD Phenom X3/X4 processors are horribly strong. The Q6600 on both the 680i SLI motherboard and this new GF9300 motherboard perform roughly equal.

Memory test

Lavalys Everest has an excellent tool to measure memory bandwidth. I don't know exactly what was going on, but the GF9300 ECS mainboard lacked an edge in memory performance. Likely something needs to be dealt with in the ECS BIOS. Manually setting a higher frequency or divider did not change a thing.

The downside here is that the overall performance will be affected due to, what I believe to be, a BIOS bug. Look closely at the read performance, that should have been 6500/7000 MB/sec.

Write performance was decent and on par with the others though.