Hitachi 7K1000.C benchmarks

[dc_v01]

I was looking at the new 1TB Hitachi 7K1000.C that was posted to the performance charts - and it looks pretty strange with the IOmeter results. In general it's quite competitive - I'm comparing it to the 7200.12 and Caviar Black. It seems to dust the competition in one benchmark - IOmeter streaming reads in MB/s - where it gets a nonsense number like 900+MB/sec vs 100, which I don't believe. But in other IOmeter tests, like database benchmark IOs/s (37 vs ~100), streaming reads IOs/s (~120 vs 900), and streaming writes IOs/s (~130 vs 900) it gets destroyed. What gives?????

 

[sub mesa]

How big test file are you using for IOmeter? You need at least 8 times your RAM capacity. So with 4GB RAM your test file should be at least 64GB. This will prevent you from actually testing your RAM instead of testing disk access; since the data you requested is still present in RAM you get high scores. By using a larger test file this will not fit into RAM so it will have to do real disk access.

 

[dc_v01]

Hi sub mesa, thanks for the reply. To be clear, I wasn't running the test - I was thinking about buying this drive, and I looked at the "2009 3.5” Desktop Hard Drive Charts" posted here on tomshardware.com. I don't know how the test was run. I can see from your description how running with an insufficiently large file size could boost that one score - but I don't see how it could account for the other scores being so dramatically low - lower performance than many previous generations of drives. Something else may not be right - and I'm worried it's the firmware on the drive!

 

[sub mesa]

Ah i can see now.

Well, 900MB/s would ofcourse not be even possible to go through the SATA cable; which limitd to 300MB/s. So at least part of the data was in RAM during that IOmeter test; so you should ignore the 900MB/s score as its bogus.

What will you be using the HDD for? If for large files, you need not to worry. If for anything else; all HDDs are slow with this; the actual differences arent that big. For a real difference when using the HDD as a system drive containing the operating system, you need a solid state drive instead.

In other words, HDD performance is not that crucial; as all disks in the same generation perform roughly the same.

 

[dc_v01]

Good point about the SATA bandwidth, that proves it for that score. I'm more worried about the low ones....

What will you be using the HDD for? If for large files, you need not to worry. If for anything else; all HDDs are slow with this; the actual differences arent that big.

I'm getting 4 for a RAID 5 in a small NAS. Small multiuser environment. (3 PCs) Was looking at these drives, but now I think I'm going for Seagate 7200.12s. The Hitachis were cheaper, and the Seagates seem to have a relatively slow access time, AAM which can't be disabled, which I thought might be bad for this. But in application benchmarks, the Seagates seem quite decent, falling down nowhere, where the Hitachis seem to have specific issues on a few tests. I found another benchmark site, harddrivebenchmark.net, which also seems to find an issue with these new drives, almost 1/2 the score of older models or the new Seagates.

Thanks for your input!