The SSD Review The Worlds Dedicated SSD Education and Review Resource | 
ATTO Disk Benchmark is perhaps one of the oldest benchmarks going and is definitely the main staple for manufacturer performance specifications. ATTO uses RAW or compressible data and, for our benchmarks, we use a set length of 256mb and test both the read and write performance of various transfer sizes ranging from 0.5 to 8192kb. Manufacturers prefer this method of testing as it deals with raw (compressible) data rather than random (includes incompressible data) which, although more realistic, results in lower performance results.
SanDisk specifications list the Extreme at 550MB/s read and 520MB/s write which seems to be right on the money as we see results of 556MB/s read and 506MB/s write. Normally, a small fluctuation can be expected and subsequent testing will show results that may be a bit higher or lower.
CRYSTAL DISK BENCHMARK VER. 3.0 X64
Crystal Disk Benchmark is used to measure read and write performance through sampling of raw (0/1 Fill/compressible) or random data which is, for the most part, incompressible. Many new SandForce Driven SSD owners who cant wait to test the performance of their SSD often grab this program and run a quick test, not realizing that they are testing with incompressible data rather than compressible data used in testing by manufacturers. We have provided compressible (oFill) results on the left with incompressible (random data) results on the right.
It is normal for the Crystal Diskmark results on the left to be a bit lower than the previous ATTO results and these tests bring random 4k write results as high as 89MB/s into the picture. These are nice scores as those low 4k results play a definite role in the visible improvement we see when migrating from a hard drive. As well, these results demonstrate the performance difference when moving highly compressible data, as compared to very incompressible data such as music, movies or photographs.
Up until recently, AS SSD was the only benchmark created specifically for SSD testing and it uses incompressible data. AS SSD, for the most part, gives us the worst case scenario in SSD transfer speeds while using SandForce Driven SSDs as they use compression in storage as discussed earlier. Many enthusiasts like to benchmark with AS SSD for their needs.
Our AS SSD results confirm that in which we see above with the Crystal DiskMark result on the right. The difference between the two can be attributed to the testing configuration of the software, whereas, Crystal DiskMark posts the high of several test passes and AS SSD combines all for an average score.
somethin’s not right…vantage score should be ~85,000…
Thats not true. All of our testing is run within the same set of circumstances and this is indicative of the results we have consistenlty received within our site test benches. I would be concerned if this SSD hit 85,000 and the highest outside of that was only 74,000. Our guide which was provided by Sandusk, in fact, showed much lower results which could have been from dirty testing or a different configuration.
This is the beauty of using several different benchmarks; different results show through in different ways, however, it in no way negates what a great drive this is.
Your TRIM is FUBAR too just like mine was. It shows you using 10.6 drives, just like I was. I also tried 10.8, 11.0 and 11.1. My guy at RWL had the same issue and my new guy at RWL also ran into the issue with a brand new build two days ago.
This drive is faster than the Intel and is the fastest SF drive on the market. I spent two weeks backtracking my steps and testing different configurations of ORAM, chipset driver and RST driver. I found several combinations that do not work correctly.
Where are you seeing this?
The tell is in your CDM 4K single command numbers. I’m replying to your email now. I just walked in the door.
Maybe the same issue as TT?
https://www.tweaktown.com/articles/4613/24nm_flash_ssd_faceoff_sandisk_extreme_retake_and_plextor_m3_pro/index1.html
I don’t compare our results to that of other reviewers, however, appreciate your sending that along. I have no reason to believe TRIM was not working and, in fact, every indication seems to say that it was and that our results are accurate.. I will be using the Sandisk and will watch performance results closeley in the next few weeks.
Hey Les, nice review! Minor correction, page one paragraph 8 “heart and sole” should be “heart and soul”. Nice to see SanDisk making such a splash on the pricing end of this market as well as holding their own on performance. As you say, it may just be a great indicator of things to come!
S.
Amended and thanks for the heads up… It was a great review to complete.
Also, “its” and “it’s” are frequently confused:
it’s is a contraction for “it is”, e.g. it is well known that Les is enthusiastic!
its is the possessive form, e.g. its impact is predicted to be huge!
I know: this appears backwards, but THAT ENGLISH!! 🙂
«The total RAW capacity of the SSD is 256GB as each module is 32GB in size, however, firmware and over provisioning needs reduce the Extreme to the advertised capacity of 240GB. Formatting reduces this further and leaves the consumer with an end capacity of 224GB»
As usual, the capacity figures are based on ignoring the distinction between gigabytes (power-of-10 units) used by storage vendors and gibibytes (power-of-2 units) used by chip vendors and operating systems.
In particular the advertised capacity is 240 gigabytes, but MS-Windows reports capacity as 224 gibibytes, which is the same. The NTFS formatting has a small overhead that does not change the outcome significantly.
The analysis should be:
* The raw capacity of the chips is 256GiB (which is 275GB), as there are 8x 32GiB flash chip modules.
* Over-provisioning accounts for around 14% of total capacity, leaving 240GB (which is 224GiB) as the visible volume capacity.
As much as we appreciate your clarification, this SSD is greared towards the consumer who would not understand such. For simplicity sake, each module has a capacity of 32GB (gigabyte) which totals 256GB. Over provisioning and firmware reduce that to the advertised 240GB. In my own opinion, the units of GiB and GB are an absolutely unnecessary distinction which only serves to cloud something that is a very simple and clear explanation.
Yes, I see: this is not a technical site (with technically correct formalism), but a consumer oriented site, with consumer-oriented lack of formalism.
no one uses GiB/KiB/MiB on any site that i am aware of. it is unnecessary for the purposes of general reviews.
Usually only people who like to argue bring this up….this sounds like someone i know ;o