Information

Dear forum members,
We are delighted to inform you that our forums are back online. Every single topic and post are now on their places and everything works just as before, only better. Welcome back!



Discussion

Discussion on Article:
Kingston SSDNow V300 Solid State Drive Review: SandForce with New Fixings

Started by: Alereon | Date 05/14/13 07:50:38 PM
Comments: 14 | Last Comment:  04/16/14 03:59:37 AM

Expand all threads | Collapse all threads

[1-4]

1. 
Regarding the quality of the flash memory, Kingston is buying whole untested wafers from SanDisk-Toshiba and doing all the rating and validation themselves. This means that the memory is only as trustworthy as Kingston's validation process. I think all indications are that Kingston drives are not reliable for this reason, but without good return rate statistics it's hard/impossible to prove.

It's generally considered that drives from NAND manufacturers (Intel, Samsung, SanDisk) are the most reliable, followed by companies that buy branded NAND directly from them, followed by companies that buy wafer-level and do their own validation, with companies buying discount/unbranded/failed NAND on the spot market coming in dead last for reliability.

For example, OCZ was known for using Spectek-branded flash memory, which is the brand Micron uses for its memory that fails validation. This was likely the primary cause of the extremely high failure rates (~5X the industry average as I recall) that gave OCZ their bad reputation. KingFast also (apparently unwittingly) sent out some review sample SSDs based on counterfeit NAND that was originally rejected by OCZ, remarked as genuine Micron NAND, and sold on the spot market.
0 0 [Posted by: Alereon  | Date: 05/14/13 07:50:38 PM]
Reply
- collapse thread

 
Interesting. I haven't heard of such claims. Can you provide references?
0 0 [Posted by: gamoniac  | Date: 05/15/13 06:14:15 AM]
Reply
 
Regarding wafer-level validation, Kingston's statement on Page 2 discusses that they buy wafer-level and do their own validation, Anandtech's V300 review has more details: http://www.anandtech.com/...ngston-ssdnow-v300-review

OCZ's use of Spectek memory is also discussed in Anandtech's review of the OCZ Vertex 3, lots of other references are also available: http://www.anandtech.com/...ocz-vertex-3-review-120gb

Anandtech also mentions that drives from NAND manufacturers are more reliable than others in their Samsung 840 review, I leave Crucial/Micron off of the list due to their history of firmware problems: http://www.anandtech.com/...he-endurance- of-tlc-nand

When they were still being published, BeHardware's component return rates showed that OCZ drives were returned at around 5X the rate of competing drives: http://www.behardware.com...ents-returns-rates-7.html

I provided mostly Anandtech links because of my very high opinions on the quality and accuracy of their SSD coverage in particular.
1 0 [Posted by: Alereon  | Date: 05/15/13 01:37:36 PM]
Reply
 
Much appreciated! I read most of their articles, too. I will check out those links.
0 0 [Posted by: gamoniac  | Date: 05/16/13 07:40:16 PM]
Reply

2. 
For all practical purposes any of the SSDs on the market will deliver essentially the same performance in actual use. No one can visually detect a difference in performance from the slowest SATA II SSD to the benchmark fastest SATA III SSD. Thus those looking for an SSD should shop for an SSD based on price, warranty terms and conditions and SSD maker reputation.

As we have seen time and time again with GPU cards, manufacturers often write algorithms designed to achieve high benchmark results at the expense of actual performance in everyday use. Thus the benches have become insignificant and often unreliable. Buy what makes you happy but don't be duped by review benches that often do not resemble the actual user experience.
2 2 [Posted by: beenthere  | Date: 05/14/13 09:10:45 PM]
Reply
- collapse thread

 
Absolutely agree. I got a few entry-level SSDNow V+100 and V+200 back when SSD were expensive. All have been on 24x7, log over 16000 hours and running 3 VM full time.

Granted, nowadays, SSD prices have come down. The price difference between high-end and entry level are only a few tens of dollars, it might be wise to grab the better ones anyway. Last I checked newegg, V300 is way more expensive than some of the better ones on XBitLabs' SSD chart (on the conclusion page). So it doesn't make sense to get V300 unless they bring the prices way down.

[Edit]: One would notice the performance difference between a good vs mediocre SSD when under heavy load, but for the most part, the difference is not very noticeable for "most" users. That said, I would rather grab a superior SSD when the price is right.
1 0 [Posted by: gamoniac  | Date: 05/15/13 06:07:44 AM]
Reply
 
"No one can visually detect a difference in performance from the slowest SATA II SSD to the benchmark fastest SATA III SSD."

This just isn't true. Crucial V4 and old JMicron drives have significant stutter issues.
0 0 [Posted by: rrr  | Date: 05/25/13 02:29:47 AM]
Reply

3. 
I have the feeling that the drop is due to 19nm MLC NAND. 25nm MLC NAND hits usually about 3k cycles (like the HyperX3K) while 34nm MLC NAND hits about 5k cycles (original HyperX)
0 0 [Posted by: Stickmansam  | Date: 05/15/13 01:37:00 PM]
Reply

4. 
Hi, I was wondering your opinion on this article.
It is true that Kingston is cheating-falsifies ?
http://diit.cz/clanek/pomaly-kingston-ssdnow-v300
(translate google or enter) http://translate.google.c...maly-kingston-ssdnow-v300

I wrote a query to the Kingston web but no reply ..
Sorry for bad English
0 0 [Posted by: Koltak  | Date: 04/16/14 03:59:37 AM]
Reply

[1-4]

Back to the Article

Add your Comment