Single page Print
Winbench starts out with some synthetic tests that measure raw performance, but then gets into some application-specific testing as well. First up is the start and ending disk transfer rate.

RAID 0 shows its impressive throughput again, doubling the scores for RAID 1 and ATA100. What's interesting to note here is that the only configuration able to sustain its throughput is ATA66, which starts and ends slowly, but at least at the same speed. ATA100 and RAID, on the other hand, lose some of their speed towards the end, which we can live with given the overall increase in performance.

Next up is Winbench's Disk Access Time test, which measures, you guessed it, disk access time. This benchmark measures physical performance characteristics of the hard drives themselves.

RAID of any flavour doesn't make a difference for a test that depends on the mechanicals of the individual drives. No matter what RAID configuration you go for, it's not going to help your drives' heads move any quicker.

CPU utilization is next in Winbench's queue. As we all know, nothing comes without a price. This test will give us an idea of what the computational costs are for each RAID configuration.

As you can see, ATA66 manages to pull this one out, though none of the values measured are unreasonably high. CPU utilization, at least as far as Winbench is concerned, is tied to the Highpoint 370 controller used by both the RAID and single ATA100 configurations.