The new firmware provides very impressive numbers and takes SSDs to a higher level of performance well beyond HDDs as system drives in desktop computers.
I'm not usually a stickler about this, but was this article translated or something? On the first page it seemed like there were a lot of grammatical issues, far more than the occasional type-o that is to be expected.
This looks very good. It's great to see that they fixed the problem that was causing SSD slowdown over time. I think I'll have to put one of these in my new build.
A new firmware for the Intel X25-M G1 is nice and all, but the SSD still fails price per performance per space against Crucial's M225 and Corsair's P128/P256. All you get with the X25-M G1 is great I/O performance. Write speeds are still pretty low compared to going up to 200MB's W/s.
The random read/writes are through the roof. The other SSD's are faaar behind. You might not reach the high throughput as the corsairs you mentioned but then it depends on what you intend the drive to do.
I have a raid for big file transfers with conventional drives, those will handle throughput, and will have SSD for smaller file operation, OS, games, etc. The X25-M G2 seems nice for that!
Is there a typo in the Read Throughput Graph? It looks like with TRIM read throughput drops after use? The labeling is not consistent with the graph below it so hopefully trim does not adversely affect read speeds...
jezza, i totally agree. I also believe that each ssd are better then others for certain environments, such as SVoyager put.
I'm not a fanboy of any of these companies, but i think intel needs to react to the competition, besides the few task where the intel ssd hase an advantage they can be easily set aside for a similar priced drive with more performance in general.
For the intel fans they can always buy two 160gb drives and raid them for a whopping 140MB/s write at a cost of just over $1300. The extra pcie raid controller they might need, might push that 1300 into 1500 territory.
ER i'm going to go out on a limb here and say "its because of IOPS". Why are people getting hung up on sequential reads and writes? Heck if you want that just go buy a Sata II drive. No spindle can touch the IOPS of any decent SLC or MLC.