I've recently swapped my HD for a 300gb VelociRaptor, seems to be running okay, but not particularly fast, I tested it against my secondary 7200 rpm hard-disk using DiskBench and it was coming out considerably slower. Write Caching is not enabled on either drive. I've also read something about IDE controllers and Ultra DMA Mode 5, though I've no idea how to identify the current setting or change them.
My Machine: HP Pavillion Core 2 Quad 6600 2.4GHz 3gbRAM 300gb VelociRaptor Windows Vista
The Velociraptor should be a bit faster for random I/O, but don't expect it to hold up against 7200rpm Terabyte drives for sequential I/O. In terms of transfer rates the increased data density of the larger drives more than makes up for the slower rotational speed.
For random I/O both types are blasted out of the water by SSDs. If capacity is not an issue then you're a lot better off spending your bucks on them.
The Velociraptor should be a bit faster for random I/O, but don't expect it to hold up against 7200rpm Terabyte drives for sequential I/O. In terms of transfer rates the increased data density of the larger drives more than makes up for the slower rotational speed.
For random I/O both types are blasted out of the water by SSDs. If capacity is not an issue then you're a lot better off spending your bucks on them.
Not true at all. The Velociraptor keeps up with and in most cases beats (though not by a very large margin) any 7200rpm drive for sequential, and it absolutely blows them out of the water for randoms.
xekuter: what kind of benchmark numbers are you getting? The velociraptor should definitely be the faster drive. Oh, and write caching should be enabled for best performance (though it shouldn't be necessary for the velociraptor to beat the 7200).
The Velociraptor keeps up with and in most cases beats (though not by a very large margin) any 7200rpm drive for sequential, and it absolutely blows them out of the water for randoms.
Hmmm - I tracked down some reviews and I see that you're correct. Shame on me for spouting off without checking my facts first...
Thanks for your help guys... CJL here's the results from DiskBench (Block size 2mb):
Created file: C:\Users\X\Desktop\DiskBench1.bin
Size: 52428800 bytes
Time: 306 ms
Transfer Rate: 163.399 MB/s
Created file: K:\DiskBench1.bin
Size: 52428800 bytes
Time: 231 ms
Transfer Rate: 216.450 MB/s
C is the Velociraptor, K is the Samsung 7200.
Will the test be affected by the fact that the O/S is running on C? (I did however wait until the machine had 'settled' and that there were no other apps runnning).
Are those RAID arrays or something? Those numbers are way fast for a single drive, to the point of impossibility. One velociraptor should be around 105MB/s sequentials, with 7ms access time. One Samsung 7200 should be around 90-100MB/s sequentials with 12-15ms access time.
------------------------------Asus P6T deluxe
i7 965 @ 4.2GHz (200*21), 1.384V
12GB Corsair Dominator DDR3-1600 CAS 7
Reply to cjl
OK, I just downloaded diskbench myself, and my WD Caviar Black scored over 300MB/s. It strikes me that the create file benchmark is probably going straight to cache (and your write caching is probably on). My single Caviar Black also beats my RAID array with 2 velociraptors on that benchmark (and does so by a fair margin), which makes it seem somewhat unrealistic to me (since I know my velociraptors are faster in every way). It also claims almost 2000MB/s for a 2GB file, which really kind of triggers the BS meter for me. Try a different benchmark, such as HDtune.
Message edited by cjl on 08-30-2009 at 11:33:12 AM
------------------------------Asus P6T deluxe
i7 965 @ 4.2GHz (200*21), 1.384V
12GB Corsair Dominator DDR3-1600 CAS 7
Reply to cjl
Sorry CJL new to all this, my drives are not RAID and there's no write caching as far as I can tell ("Enable Write Caching on the Disk" not checked). I'll download HDtune now and report back!
CJL Thanks for all your help by the way, here's the results from HD Tune, slightly more encouraging, though I was expecting a more dramatic difference:
HD Tune: WDC WD3000GLFS-01F8U Benchmark
Transfer Rate Minimum : 72.7 MB/sec
Transfer Rate Maximum : 118.4 MB/sec
Transfer Rate Average : 99.4 MB/sec
Access Time : 7.3 ms
Burst Rate : 158.5 MB/sec
CPU Usage : 2.3%
HD Tune: SAMSUNG HD753LJ Benchmark
Transfer Rate Minimum : 61.5 MB/sec
Transfer Rate Maximum : 113.8 MB/sec
Transfer Rate Average : 90.0 MB/sec
Access Time : 14.9 ms
Burst Rate : 135.9 MB/sec
CPU Usage : 2.0%
Those look about right. By the way, the huge difference is not the sequentials (though you can see they're around 10% faster), it's that access time. That means that every time some tiny little file needs to get accessed on the samsung, it takes 15ms for the drive to even get to where the file is, before it can even start reading it. On the velociraptor, it only takes 7ms. That's a huge difference for things like boot times. That's also the reason SSDs feel so fast - they have extremely short access times.
------------------------------Asus P6T deluxe
i7 965 @ 4.2GHz (200*21), 1.384V
12GB Corsair Dominator DDR3-1600 CAS 7
Reply to cjl
TBH the drive is probably wasted on me, I'm sure it could be 'tuned' to do so much more, but I'd had 12 months of running a dodgy Seagate drive thinking it was software/OS and when I finally discovered it was down to the drive I wanted to replace it with something solid... definitely feels faster though wanted to make sure it was running as it should be!