backtop

Hardware Kingston Launches New SSDNow V+ Drives
Shane McGlaun (Blog) - August 11, 2009 1:13 PM
Print E-mail del.icio.us 25 comment(s) - last by deegee.. on Aug 14 at 3:40 PM
Drives are offered in up to 256GB for a cool $992

SSDs are creeping further into the mainstream as manufacturers introduce more SSDs onto the market and prices are driven down. Many memory makers have started producing SSDs as well including OCZ and Kingston among others.

Kingston has announced its latest SSD called the SSDNow V+. The new SSD will be offered in capacities up to 256GB and is designed for higher performance than comparable hard drives of the same capacity. Kingston will also offer the SSD in 128GB and 64GB capacities. The 256GB SSD will retail for $992, the 128GB will sell for $500, and the 64GB will sell for $254.

Like all SSDs, the SSDNow V+ is designed for performance. Sequential write throughput for the 64GB drive is 140MB/sec, the 128GB SSD is good for 170MB/sec, and the 256GB SSD is good for 180MB/s. IOPS  with random 4K files for read in 6,300 IOPS, random 4K writes for the 64GB is 84 IOPS, the 128GB is 158 IOPS, and the 256GB is 291 IOPS.

The SSDs all use the same 2.5-inch form factor and connect via either SATA 1.5Gb/sec or 3Gb/sec interfaces. SSDNow V+ drives measure 69.85mm x 100mm x 9.5mm and weighs 84 grams. Operating tolerance for the drive is 2.7G, non-operating tolerance is 20G, and the drives can withstand an operating shock of 1500G. Life expectancy for the entire line is 1 million hours MTBF. Power consumption is 2.6W in active mode and 0.15W at idle.

"The SSDNow V+ is the ideal solid-state drive for power users, system builders, system integrators, and in demanding corporate environments where efficiency and performance are important as more system resources are in use," said Ariel Perez, SSD business manager, Kingston(R). "Users will be more productive as the drive's higher capacities and IOPS allow for more data storage and faster boot and application load times."

Comments     Threshold


This article is over a month old, voting and posting comments is disabled
Random Read/Write
By lotharamious on 8/11/2009 2:47:51 PM , Rating: 2
Knowing IOPS is great and all... but what about the data rates of its random reads and writes?




RE: Random Read/Write
By mckirkus on 8/11/2009 3:30:48 PM , Rating: 4
IOPS are data rate. 291 4K IOps/S == 4kB * 291 = 1.1MB/s.

This think has pretty mediocre 4k write IOPS. I think the Intel drives are doing 20+MB/s, even the SLCs.

They don't talk about the sequential read but it looks like about 250MB/s if you do the math on the read IOPS.


RE: Random Read/Write
By mckirkus on 8/11/2009 3:31:05 PM , Rating: 2
IOPS are data rate. 291 4K IOps/S == 4kB * 291 = 1.1MB/s.

This think has pretty mediocre 4k write IOPS. I think the Intel drives are doing 20+MB/s, even the SLCs.

They don't talk about the sequential read but it looks like about 250MB/s if you do the math on the read IOPS.


RE: Random Read/Write
By Mr Alpha on 8/11/2009 4:11:50 PM , Rating: 2
1.1MB/s 4k random write isn't mediocre, it is lousy. Intel's new 34nm MLC drives do 34.5MB/s. Even the OCZ Vertex gets above 10MB/S.

Wonder if this Kingston SSD uses a Samsung controller?


RE: Random Read/Write
By tastyratz on 8/11/2009 4:16:42 PM , Rating: 4
Makes you wonder
Seems like substantially more money for lower overall performance over their competitors. This drive hardly has me wooed...
in a world where new ssd drives are flavors of the week, I say:
NEEEEEEEEXXT


RE: Random Read/Write
By Bigjee on 8/11/2009 4:19:17 PM , Rating: 2
These drives use a Jmicron controller with a 64K cache.

And the article claims the 64gb is for $250?
You can get it at new egg for a $US 134.

http://www.newegg.com/Product/Product.aspx?Item=N8...


RE: Random Read/Write
By amanojaku on 8/11/2009 6:07:38 PM , Rating: 2
quote:
And the article claims the 64gb is for $250? You can get it at new egg for a $US 134. http://www.newegg.com/Product/Product.aspx?Item=N8...
Sorry, but you're incorrect. That is one of the old "V" models, which are being replaced by the "V+" described in this article.

Old:
SNV125-S2/64GB
SNV125-S2/128GB
SNV125-S2BN/64GB
SNV125-S2BN/128GB
SNV125-S2BD/64GB
SNV125-S2BD/128GB
http://www.kingston.com/ssd/V-series.asp

New:
SNV225-S2/64GB
SNV225-S2/128GB
SNV225-S2/256GB
http://www.kingston.com/press/2009/flash/08a.asp


RE: Random Read/Write
By deegee on 8/11/2009 5:23:21 PM , Rating: 1
-=> wall of text warning <=-

90%+ of the average desktop use for user and system read/writes are not 4k random, so we need to keep things in balance here. Too many review sites are only posting 4k rnd and 2MB seq, and that tells us nothing of the drive's actual performance curve.

Hopefully there is going to be a lot of the myths regarding SSDs pushed aside soon, but you never know with consumers. Things such as shutting off almost everything in Windows is pure bunk (page file, indexing, etc.), unnecessary, and goes against the entire reasoning for having a faster drive. Read this: http://blogs.msdn.com/e7/archive/2009/05/05/suppor...

Some manufacturers claim that you should disable as many write-oriented features as possible to extend life, but try calculating lifespan to see what the reality is.
Assuming an 80GB SSD with a Write Endurance of 10,000 per cell, and you write 2MB of data every second continually 24:7:365 (which is 172.8GB every day!), then the expected lifespan of the drive is 400,000,000 seconds or 12.68 years (10,000 * 80,000,000,000 / 2,000,000).
Even if you dropped cell endurance in half assuming additional erase/writes caused by block updates, that results in 6.34 years. Now let's drop our usage down to only 12 hours and 86.4GB a day (how many desktop users actually write that much each day?), and we are still back at 12.68 years lifespan. I know that I don't write anywhere near 86GB each day to my drives...

Regarding the new Intel 34nm MLCs, their 4k is really high, but they suffer badly on larger read/writes behind almost every other "performance" drive, which in the real-world is going to drop their entire performance curve much lower. Personally I want to see some actual in-use reports before I agree that their performance is going to be good according to benchmarks alone.


RE: Random Read/Write
By xrodney on 8/12/2009 4:49:57 AM , Rating: 2
I hope you do realize that your 12.68 years is time after which there will be no single cell working ?


RE: Random Read/Write
By deegee on 8/12/2009 3:03:55 PM , Rating: 2
Working? No, writing . The drive can still be read, the cells just stop erasing/writing.
How many people here write 172GB every day to their drive and are still using the exact same magnetic hard drives in their computers after 12 years?
If you wrote 17GB every day you would still get 120 years before the cells stopped accepting writes.


RE: Random Read/Write
By Mr Alpha on 8/12/2009 5:39:08 AM , Rating: 2
The sequential read of the Intel drive is 250MB/s which is up there with the other performance SSDs, and getting close to the real world max sustained transfer rate of SATA 3Gbit/s, which is somewhere at 260-270MB(s. The sequential write of 80MB/s of the Intel SSDs is still their weak spot. But I wonder how often this is a problem. How often do you move big chunck of data on an SSD where the sequential write speed would be the bottleneck? If you copy from another SSD, then sure. But even when copying big files from a hard drive; most hard drives don't have average sequential read speeds much above this.

The 4K random, and the sequential tests as well, are all synthetic and I like real world benchmarks as much as the next guy. But when looking at synthetic benchmarks I'd argue that the 4K random is more important than the sequentials, even if real desktop usage only to a small part consists of 4K random, because the sequential performance matters almost never. You notice this if you stick a bunch of hard drives in RAID 0, because it increases the sequential performance, but you don't notice any real performance improvement.


RE: Random Read/Write
By B3an on 8/12/2009 6:22:53 AM , Rating: 2
quote:
But when looking at synthetic benchmarks I'd argue that the 4K random is more important than the sequentials


You're right, they are way more important. AnandTech points this out aswell. And from owning two of the new 34nm Intel SSD's i can also back this up.

They are faster than my previous OCZ Vertex SSD's (still great drives) which had considerably higher sustained write. The only single thing where the Vertex's are faster is when copying a single large file.

Too many people look at sustained read/write performance when looking at SSD.s. The Random read/write matters a lot more.


RE: Random Read/Write
By deegee on 8/12/2009 3:40:05 PM , Rating: 2
I read the Anandtech reviews and reports, however I don't rely on any of their tests. And I don't mean that as a flame or slam on them, just my opinion, so no reason for the fanboi's to start flaming me. :-)
I'm not a newb either, I've been in the computer field for almost 30 years, doing hardware/PCB/IC design, embedded systems, robotics, ucomp assembler os's, etc.

As an example, look at AAT's Intel X25-M G2 Preview review here:
http://www.anandtech.com/cpuchipsets/showdoc.aspx?...
(I don't understand why they have this listed under CPU/Chipsets either...)

And compare their "thin" 4k and 2MB results information with someone such as one of FLC's reviews on the Kingston SSDnow V Series here:
http://www.futurelooks.com/kingston-ssdnow-v-serie...

By providing a full performance curve, such as the ATTO graphs shown on FLC reviews, it's easy to see significant differences in total performance that you simply cannot get from just the 4k and 2MB test that AAT relies on.


RE: Random Read/Write
By Howard on 8/12/2009 11:21:08 PM , Rating: 2
Could you explain how your experience is relevant to the field of large long-term data storage?


RE: Random Read/Write
By deegee on 8/12/2009 3:14:23 PM , Rating: 3
"The sequential write of 80MB/s of the Intel SSDs is still their weak spot. But I wonder how often this is a problem. How often do you move big chunck of data on an SSD where the sequential write speed would be the bottleneck?"

It still places them behind the other drive manufacturers is my point.
And Sequential Write doesn't just mean 1GB video files. The majority of writes are sequential, from most user data files to the paging file.
Sites need to start showing the full performance curve in their reviews/tests, some sites do and it is really informative. If 4k Rnd and 2MB+ Seq were artificially high on a drive, and its 32k through 1MB Seq performance was poor, then overall performance will be poor since that is where most writes occur at.

I have never stated that 4k rnd doesn't matter at all, it just needs to be kept in perspective of real-world use. Many of the review sites and even benchmark software target only 4k and 2MB sizes, which tells nothing of the full drive performance curve. And you can't make assumptions as to the other performance values based on just these two.


RE: Random Read/Write
By deegee on 8/14/2009 3:40:45 PM , Rating: 2
I should mention that my calculations are intentionally not including Write Amplification, premature cell death, and erase block size. Mainly because these specs are rarely mentioned by SSD manufacturers.
Worst-case death can occur as soon as less than one year under the proper conditions. However, the average desktop user should get at least 5 years out of a 64GB SSD. I've been pushing first-gen controller SSDs under heavy desktop use just to see what the results are.


What I want to know...
By Boze on 8/11/2009 4:09:56 PM , Rating: 3
...is why we haven't seen an AnandTech review on FusionIO's upcoming ioXtreme PCIe drive. I'm currently tossing around the idea of using it for my next computer, since it seems to be vastly ahead of anything any other company is offering, and while $895 for 80 GB of MLC storage isn't exactly stellar, you definitely can't argue with the performance.

Or maybe I'm jumping the gun. Maybe in another year, SATA 600 MB/s internal connections will be available on motherboards, extremely fast SLC drives will have come to prominence, and all for a fraction of that $895.




RE: What I want to know...
By Fnoob on 8/12/2009 12:22:47 AM , Rating: 2
You must be dreaming up a box to run Crysis, right?


RE: What I want to know...
By xrodney on 8/12/2009 4:44:35 AM , Rating: 2
For 895$ its possible get Fusion IO drive, only 80GB capacity but speed over 500MB/s (more then enough for OS and few games).
Was lately looking to get high performance SSD, but i am still more and more inclining to get fusion IO as soon as they release FW update for supporting booting.

Review here:
http://www.dvnation.com/Fusion-IO-IODrive-SSD-Soli...


Are these SLC drives?
By Jedi2155 on 8/11/2009 2:31:38 PM , Rating: 2
At 170 MB/sec write performance, these sound like SLC equipped drives. A 64 GB SLC drive @ $3.90/GB is a nice drop from the $10.92/GB Intel is charging for its X25-E drives.




RE: Are these SLC drives?
By DOSGuy on 8/11/2009 2:49:47 PM , Rating: 2
A 256 GB SLC? In a 2.5" form factor? We're not there yet. Intel's G2 drives only use one side of the PCB, so they could achieve capacities of 320 GB MLC or 160 SLC using a state-of-the-art 34 nm NAND chips. We're still a generation away from 2.5" 256 GB SLC drives.

Of course, you have followed the link to the Reviews section and read "The Kingston SSDNow V Series really does offer great value for those looking to move to an entry level MLC SSD without breaking the bank."


1 million MTBF
By scrapsma54 on 8/12/2009 4:04:40 AM , Rating: 2
Thats considerably better than pretty much every known SSD on the market. This is the first good technologically and non-political news I have heard from DT in a while.




RE: 1 million MTBF
By dhalilahma on 8/12/2009 4:15:52 AM , Rating: 2
Meh
Google - press releases
click - ctrl-C
click - ctrl-v
Shazaam, you are a journo
?
profit


Garbage Collection?
By Zshazz on 8/11/2009 4:00:44 PM , Rating: 2
This is cool and all... but what I think is more important is if this drive will have any garbage collection enabled. It's a piece of information that I don't see on their site, so I'd assume not. Anyone know for sure?




Write performance
By x0rg on 8/12/2009 7:16:24 PM , Rating: 2
For Intel X25-M G2 drives write IOPS are:
8600 (for 160GB version), 6,600 (for 80GB version).
Read IOPS - 35,000 (Kingston offers 6300 IOPS).
Now, compare these numbers with those Kingston offers.
84, 158, 291 IOPS??? Is that regular hard drive or SSD?
Is anybody going to buy that?




"There's no chance that the iPhone is going to get any significant market share. No chance." -- Microsoft CEO Steve Ballmer









botimage
Copyright 2010 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki