try those tests again with write cache disabled and let me know what you get.
http://www.tomshardware.com/charts/3-5-hard-drive-charts/average-read-transfer-performance,658.html?p=2024%2C1793%2C1832%2C1783%2C1794%2C1796%2C1808%2C1779%2C1833%2C1828%2C1806%2C1791%2C1811%2C1803%2C1767%2C1819%2C1789%2C2023%2C1840%2C1837%2C2021%2C1843The fastest read speed off a 7200 RPM drive is 100 megs per second and that's a hug exception to the general trend. 95% of 7200 RPM drives will no top 80MB/sec average. My 15,000 RPM SCSI 74GB drive reads at 92 megs per second. I highly doubt you're going to top that.
According to that chart above, the WD3200 doesn't even top 55 megs per second average.
I'd love to see you try to benchmark that drive with ATTO and see what you get. Then we'll talk about true hard drive performance.
Oh, and in regard to what you noticed using RAID 0, you had it all wrong if you thought RAID 0 will give you a performance benefit in day-to-day applications. RAID 0 write times are much lower than standard write times, and since all applications you use will both read and write (some of which will attempt to do both at the same time), your performance will suffer. RAID 0 is only logical when you're constantly reading very big files, such as when using photoshop or a video editing application with raw AVI files. In just about any other applications, the read benchmarks will be quite nice, but the actual performance with applications might even be lower than when using a single drive. You also have to consider that the RAID operations will also use up CPU power, which will further slow down whatever application it is that you're using.
To summarize, don't run a RAID 0 just because the benchmarks look nice.There's more to performance in data storage than just raw read-only speeds. This is why I use a $750 SCSI RAID controller with a backup battery and 128MB of dedicated ECC DDR RAM for my RAID 0 array.
Author
Topic: SCSISCSISCSISCSISCSI (Read 435 times)