I downloaded UnixBench 5.1.3 and ran it on both my old Dell 4600 and my newer Gateway SX2800. With a 2.4Ghz P4 the Dell gets clobbered on cpu tests by the Gateway with quad core Q8200 at 2.3GHz.
The surprising thing is that the old Dell with a Maxtor STM3200820A (ATA) drive is much faster than the newer Gateway with a WDC WD6400AAKS-22A7B2 (SATA) on file tests. For example:
File Copy 1024 bufsize 2000 maxblocks:
That was the "4 CPUs in system; running 1 parallel copy of tests" result for the Gateway.
Based on UnixBench lookups on the web, the Gateway result looks really low, rather than the Dell being wrong/high.
Here is my full hdparm result on the Gateway:
UPDATE: new hdparm results and full UnixBench
first of all, I did an hdparm -Tt and results were more in line with what I'd expect for the two systems:
But here are the full results, first the old Dell:
And then the new Gateway:
Update2: I found this on the web. Fact or fiction?
I switched from the 32-bit PAE kernel to a 64-bit kernel (Ubuntu 13.04), and disk writes are now fast.