WD, Intel, Corsair, Kingston, Plextor SSDs Collide 56
J. Dzhugashvili writes "New SSDs just keep coming out from all corners of the market, and keeping track of all of them isn't the easiest job in the world. Good thing SSD roundups pop up every once in a while. This time, Western Digital's recently launched SiliconEdge Blue solid-state drive has been compared against new entrants from Corsair, Kingston, and Plextor. The newcomers faced off against not just each other, but also Intel's famous X25-M G2, WD's new VelociRaptor VR200M mechanical hard drive, and a plain-old WD Caviar Black 2TB thrown in for good measure. Who came out on top? Priced at about the same level, the WD and Plextor drives each seem to have deal-breaking performance weaknesses. The Kingston drive is more affordable than the rest, but it yielded poor IOMeter results. In the end, the winner appeared to be Corsair's Nova V128, which had similar all-around performance as Intel's 160GB X25-M G2 but with a slightly lower capacity and a more attractive price." Thanks to that summary, you might not need to wade through all 10 of the pages into which the linked article's been split.
Can we really trust reviews of SSDs? (Score:2, Interesting)
These drives are all built in Taiwan and China. The same factory may produce drives for multiple vendors. The only difference between one company's drives and another is the label on the outside of the case.
Add to that that the drives are manufactured in batches, so the quality differs from batch to batch. Also, a single vendor may use multiple factories, so drive quality may differ due to that as well.
I wouldn't put any faith in any review. The only thing you need to know is the price and return policy. Everything else is too variable to determine without a very broad review of multiple drives from a series of different batches.
Re:Can we really trust reviews of SSDs? (Score:5, Interesting)
SSD's are not commodity products where one manufacturer simply slaps a label on an OEM product. There is actually a large amount of complexity in these drives with different storage controllers, caches and memory playing a large roll in performance and reliability.
Re:Can we really trust reviews of SSDs? (Score:2, Interesting)
Actually, after reading the summary and skimming through TFA, I'm inclined to say that whole review is bunk. I know for a fact that Kingston's SSDnow series are rebadged Intels, so they should perform identically. The main difference is Kingston's versions tend to be priced more aggressively.
For my money, it's either OCZ Vertex, Intel X25 (or Kingston), or nothing at all. Having tried many of the off-brand ones and been burned, I stick to what I know works. A lot of the cheaper SSDs act like someone duct-taped a SATA bridge to a USB flash drive: high latency, short lifespan, writes 4 times slower than reads... It's all about the controller chip's performance, which all but a handful suck ass.
Better article on anandtech (Score:5, Interesting)
Better because it examines the performance of a new controller series from SandForce which beats the performance of these ones by using lossless compression to write less data.
http://www.anandtech.com/print/3656 [anandtech.com]
(printed view has no ads and no margins and is one big long page...)
They ignored Intel's driver (Score:5, Interesting)
And still the Intel drive did reasonably well. Including being 4 times as fast in the 512b random write test...
Far from complete! (Score:2, Interesting)
The review omitted perhaps the most exciting SSD available to date:
Crucial C300 (6 Gbps SATA III) in capacities of 128GB and 256GB, at very competitive prices ($700/256GB on amazon).
Why no comparison with a baterry backed RAID? (Score:2, Interesting)
So I see these benchmarks between expensive SSDs and cheaper harddrives, but I'm yet to see a benchmark between some more appropriate price configurations: SDDs versus mechanical harddrives in RAID with battery-backed NVRAM, where the random write penalty is much much lower. Does anyone know of any?
iustin