I've never seen 35 on a 4k read on any drive. Thats definitely impressive. I get 25 on my 830.
I'm not saying CDM reports it correctly, but Anvil's Storage Utilities reports very similar results to CDM for 4K random read, with compressible and incompressible data.That 35MB/s from CDM for 4KiB random read is seen for many Sandforce 2281 SSDs, but it is a bogus number. If you run AS-SSD or IOMeter (set to full random), the true number is observed, and it is around 25MB/s. The higher number from CDM is an artifact of how CDM measures.
I'm not saying he is wrong but its very hard to gauge the "feel" of it. I don't think he is a newb at all but it's difficult for others since they can't feel this and benchmarks don't show a 2x improvement. There may be other things are play here and yes I know it was a drive image so there shouldn't really be any differences when it comes to the OS.
I would love to get my hands on one of these drives but I will be staying on my SATA2 board for awhile longer plus i'm adding a second intel 160 G2 so I can go Raid 0 when the 11.5 RST are released.
I've never seen 35 on a 4k read on any drive. Thats definitely impressive. I get 25 on my 830.
Although if you look at the two drives on bench, the 830 actually wins more tests than it loses (without using pen and paper) and many of the scores are very close so I do find it hard to believe there is a 2 fold in noticeable speed.
Schoolboy error on my part for forgetting about CDM being compressible.Of course there's not. Even with Intel's validation, the only thing you might see 2-fold is BSOD reports. This is Sandforce (the shadiest of the modern controllers) with Intel (the most reputable of modern manufacturers), and I'm not convinced Intel has powered through the dungpile that is Sandforce yet. Definitely going to be interesting in the next 6-12 months.
Those ssd's don't degrade very much even without TRIM, and if you somehow got into a bind it's already possible to download the beta 11.5 RST, anyway.
BTW, I read about the g2's on Anand's original writeup about them, but I've also experienced it firsthand with my x25m g2 80gb in raid 0 (SRT cache + data) for the past couple of months. Or, to put it better, I HAVEN'T experienced the need for trim on my g2 in past couple of months.
Ah interesting with that being said I may install the second ssd sooner its just sitting on my desk.
I didn't know a beta was available.
That 35MB/s from CDM for 4KiB random read is seen for many Sandforce 2281 SSDs, but it is a bogus number. If you run AS-SSD or IOMeter (set to full random), the true number is observed, and it is around 25MB/s. The higher number from CDM is an artifact of how CDM measures.
...how did you get 300MB/s for 4k random write?