Kingston SSD: Response from Tech support regarding speed.

justin4pack

Senior member
Jan 21, 2012
521
6
81
So I email Kingston support regarding the slow read/write speeds of there v300 series after reading the Anandtech review page. http://anandtech.com/show/7763/an-update-to-kingston-ssdnow-v300-a-switch-to-slower-micron-nand

This was the response I got from support:

The reason why we want you to run ATTO is because some bench marking software i.e., CrystalDiskMark, AS SSD will not provide accurate information for Sandforce based drives due to the drive's getting a lot of it's speed from compressing data. Crystal Disk and AS SSD uses incompressed data during its testing. The data cannot be compressed further so much without the write speed being affected. Because both of these benchmarks use fully incompressible data they are nowhere near a real world workload for typical desktop-notebook computing.

ATTO disk benchmark uses uncompressed data. So the Sandforce drives can perform its compression and thus the increased speed of this drive is reflected in the benchmark. It is also the program used by our engineering staff to document the speed ratings for the these drives during testing.

If you have any other questions or concerns, please feel free to reply to this e-mail with full email history. Thank you for using Kingston on-line technical support.


Regards,

Julio C Andrade
Kingston Technology
Technical Support



Is this true or a marking ploy?
 
Last edited:

Gigantopithecus

Diamond Member
Dec 14, 2004
7,664
0
71
BS marketing ploy. All Sandforce drives are marketed with ATTO numbers because those numbers look best for them. The problem, rather the BS, is the claim that other benchmark tools present a "nowhere near a real world workload for typical desktop computing." I regularly manipulate/work with incompressible data.

Kingston is rightfully catching a lot of flak for their bait and switch on these drives. I have been buying their products, from RAM to flash drives to SSDs, for years, have put them in customers' systems, and recommend their products to others. I frequently put them in the buyers' guides that I wrote for the main AnandTech site. That stopped with their unbelievably dishonest handling of these SSDs. The performance changed so dramatically, they absolutely should've said it was a different product.

If you can, you should return the drive and get something better.
 

razel

Platinum Member
May 14, 2002
2,337
90
101
It's true. If you want highest benchmark speed from Sandforce you'll want to run a benchmark with compressible data. BTW compressing data isn't a bad thing for NAND storage and is actually very smart. You can mate the right algorithm with the right processor instructions and you won't lose performance. In most cases you gain.
 

postmortemIA

Diamond Member
Jul 11, 2006
7,721
40
91
what a BS. hard disk makers could advertise burst speed of the cache as a speed for their hard drives iwth the same "logic".
The game data, movies, pictures, etc. are already compressed and cannot be compressed anymore.
 

Deders

Platinum Member
Oct 14, 2012
2,401
1
91
So In real world gaming, is there a speed difference?

Games don't use the disk very much or very quickly so I doubt you'd see much or a real world difference.

Once the game is written to the disk, the most disk activity will be game loads which in my experience rarely reaches 100MB/s, usually for less than a second.

Apart form that you might get minor read and writes whilst in game but either SSD will give you the same performance. Not that I'm on Kingston's side here.

It is a bit cheeky of them to switch like that.
 

Cerb

Elite Member
Aug 26, 2000
17,484
33
86
It's true. If you want highest benchmark speed from Sandforce you'll want to run a benchmark with compressible data. BTW compressing data isn't a bad thing for NAND storage and is actually very smart. You can mate the right algorithm with the right processor instructions and you won't lose performance. In most cases you gain.
If you want a fair test, you'll want to run a test with pseudo-random number strings as at least a fair portion of your data, if not all of it, or use files representative of real-world uses. With high quality NAND, SF controllers do not fair poorly in such tests. Look at Kingston's HyperX, Intel's 520 and 530, and Toshiba's THNSNF, for examples of such.
 

Turab

Member
Dec 16, 2013
43
0
61
With high quality NAND, SF controllers do not fair poorly in such tests.

That's right.

SF based drives are still a good option for home users.

BTW, Kristian had said that Kingston is considering to update the name into V305, is there any news about it ?