Understanding and Choosing the Best SSD

Earlier this year, Anand Lal Shimpi posted an article at his site called, “The SSD Anthology”. It was a ridiculously comprehensive look at the current (at the time) state of solid-state disks, from their perks to the leading issues. It was also interesting because it was the first time an article was published that caused multiple friends to e-mail me to tell me about it. I guess it’s no surprise that it was read over 2 million times.

Since that article was published, a lot has happened, not surprisingly. “Indilinx” is a name out of everyone’s mouth nowadays, hatred is shown towards Samsung, Intel is still reigning supreme and JMicron… who? So it was time for Anand to follow-up to his successful article, and it also happens to be a great read. I admit I didn’t read the original, but the second one was hard to put down (not that I picked my monitor up).

There’s a lot to glean from this article, and since we are soon going to be publishing SSD content of our own (stay tuned!), it was nice to be able to catch up with the SSD scheme of things. Interestingly enough, even though Intel’s X25-M G2 happens to have a weak sequential write speed compared to, well, everything else, it still comes out on top for various reasons. I find it actually rather incredible that Intel’s first product was so good, that even more than a year later, others still haven’t been able to truly compete with the overall performance.

A few points from the article that I found really interesting was that OCZ’s upcoming Solid 2 series would be utilizing Intel’s 34nm flash chips, but production hasn’t yet began because apparently those chips have slight issues with the Indilinx controller. Hopefully we’ll see those to market soon. Also interesting is the OCZ Vertex Turbo… an SSD with an overclocked controller (which makes a difference, but doesn’t scale with the price premium).

One major issue tackled is Write Amplification, which is explained in some detail in the article, and really highlights why companies need to create smarter controllers… and why TRIM couldn’t get here any sooner. If you have a few hours, the article is definitely worth the read. Thanks to On_Wisconsin for the heads-up on the article.

Credit: AnandTech

The write amplification factor is the amount of data the SSD controller as to write in relation to the amount of data that the host controller wants to write. A write amplification factor of 1 is perfect, it means you wanted to write 1MB and the SSD’s controller wrote 1MB. A write amplification factor greater than 1 isn’t desirable, but an unfortunate fact of life. The higher your write amplification, the quicker your drive will die and the lower its performance will be. Write amplification, bad.