Bought the Samsung 128gb for about 170$ 4-5 months ago and now its available for 100$. I dont think the premium of the ADATA 910 drive is worth it considering its possible to get the ADATA 900 drive for 100$ as well as the Samsung drive. One is paying a premium essentially for the extra 7% space.

Even retesting some drives because of the new controller. Good job!

The related samples in the retested benchmarks all were HDDs or Sandforce based SDDs if i've read correctly. It will be interesting to see how different controllers (Intel, Samsung and Marvel) deliver on the SPCR test suite. So i hope some other companys take this hint and sent in some review samples to take on the silent competition.

Does it squeal? Does it get hot after intense use?

Am I the only one who finds this a bit... odd? SSD are already inherently silent.

The bolded sentence doesn't make any sense. You are measuring reading, not writing speed, hence there is no wiping. The reason that the drive is slower at the beginning is because there IS some data.
So the results are:
~230MB/s reading of actual data.
~350MB/s reading of NOTHING - this is not useful at all, who would read empty space from drive?!
To sum up: average speed measured this way is IRRELEVANT.





1. Firstly, the average speed doesn't matter measured this way. Secondly, this is absolutely NOT indicative of what to expect after using drive for some time! Is is the opposite, actually. The graph show the speed of drive after formatting, who the hell uses drive in this state? The real performance of drive is drive filled with different data.
2. And what's up with that drop out to ~119MB/s?! This is very strange, and you just omit it in description, unbelievable.
3. You measure performance after TRIMming the whole drive, but don't mention how much time was given to let garbage collector recover from dirty space. That mentioned drop could disappear if you waited few minutes more. The speed and characteristics of GC are very important matter in SSD, and you just ignore it completely.


What this graph shows? Only the performance of drive just after format, without any meaningful data read actually.

The actual performance of filled drive probably will be ~230MB/s as shown in the first HD tune benchmark, at the beginning of space.

And whats up with the difference in max read speeds? New drive manages ~350MB/s max speed, but after format and TRIM its just 280MB/s. It looks like TRIM doesn't restore original performance, and once used, the drive is much slower. Again, now word about that from you.

I just registered to write this. I love your site when it comes to silence testing, but you SSD review are simply terrible. Please, at least know the matter before you write something.

The link posted to the article in question at the top is broken. Should be: http://www.silentpcreview.com/ADATA_XPG_SX910_SSD


Sometimes they make odd electrical noises which are quite noticeable.

Would be nice if there were more info on how secure this drive is. Hardly any manufacturers talk about it.

Sandforce SSDs encrypt the data, but the problem is controlling the encryption key. Where is it stored? Is the ATA password used to generate it, or hashed in some way and combined with it?

I don't have a Sandforce drive but I think they use encryption without any password in order to allow data to be wiped without overwriting it (faster, makes for a longer-lasting drive).

I think the general recommendation if you want your whole drive to be encrypted securely would be to use well-reviewed software, not hardware.
In that case, you probably wouldn't want to use a Sandforce drive because all your data would be incompressible. The software would probably initialize the whole drive with incompressible data actually.

My thoughts exactly, I mean there are so many other products I would rather see an review of at silentpcreview than SSDs (which must be the least noisy unit in a pc, electrical noise or not)...