CSX 128GB SSD
Posted 04/15/09 at 06:20:35 PM by Nathan Edwards

As the price of NAND flash memory drops to record lows, more and more hardware vendors are getting into the solid state drive business—and why wouldn’t they? A standard hard drive has lots of moving parts, but a solid state drive is nothing more than a few NAND flash modules, a controller chip, some PCB, and an enclosure. CSX is well known in Europe as a producer of aftermarket RAM for Apple products, and its Diablo gaming RAM has started making waves in the United States. But this 128GB multilevel cell (MLC) SSD marks the company’s first foray into the solid state market.
Single-level cell (SLC) SSDs typically have better write speeds than multilevel cell drives, but MLCs are more common because they’re much cheaper. We’ve tested a few standout MLCs, including Intel’s X-25M, but most of the multilevel cell drives we’ve benchmarked have suffered from poor write speeds.
The CSX SSD’s read performance of 114MB/s is second only to that of Intel’s X25-M, which clocked in at 206.6MB/s. However, write performance, at 25.58MB/s, is only slightly better than that of bottom-of-the-barrel drives from RiData and Super Talent (reviewed as part of our SSD roundup, http://tinyurl.com/9qvwyd). It’s no match for SLC drives like Memoright’s MR25.2-032/64S (http://tinyurl.com/9jnjqg), but, more importantly, it doesn’t come close to reaching the 100MB/s write time the company’s literature claims the drive can achieve.
The CSX garnered the worst Premiere Pro score of any SSD we’ve tested and a PCMark Vantage score that puts it firmly in the middle of the pack—it’s better than RiData’s Ultra-S Plus and Super Talent’s MasterDrive DX, worse than the other drives in our roundup, but not even close to our reigning champion: the Intel X-25. Large-file (greater than 504MB) random-access write times were similarly poor at 107.7ms, with max latency at nearly a full second and average large-file latency hitting 218.8ms. The Intel X25-M, by contrast, has an average random-access write speed of just .09ms.
A peek under the hood of this drive revealed the same JMicron JMF602 controller that’s used in other write-crippled drives, such as Super Talent’s MasterDrive DX and OCZ’s Core. Most SSD manufacturers have stopped using this controller; we hope CSX will do the same in future iterations of the drive.
Takes second place in read performance. Decent price/capacity ratio.
Unacceptable latency cripples write performance.
| CSX 128GB SSD | Intel X25-M | Western Digital Velociraptor | |
|---|---|---|---|
| Capacity | 128GB | 80GB | 300GB |
| Average Sustained Transfer Rate Read (MB/s) | 114.56 | 206.65 | 98.31 |
| Average Sustained Transfer Rate Write (MB/s) | 25.58 | 64.30 | 98.22 |
| Random Access Read (ms) | 0.18 | 0.12 | 7.24 |
| Random Access Write (ms) | 107.7 | 0.09 | 3.42 |
| Premiere Pro (sec) | 708 | 621 | 383 |
Best scores are bolded. All drives were tested on the same machine; a stock-clocked Intel Q6700 on an EVGA 680i SLI board with SATA II connections, using h2benchw, HD Tach Raw 3.0.1.0, Premiere Pro CS3, and PCMark Vantage 2005.
To claify, you say," If we
Submitted by rob41 on Thu, 04/16/2009 - 11:22am
To claify, you say," If we haven't tested something with our own benchmarks on our
own rigs, it doesn't exist for us." That excuse couldn't be more lame Nathan. It's not like the multitudes of faster SSD's are a rare item that's hard to get. So keep your head burried in the sand and tell people that your world is the only one that matters. If you haven't tested it it doesn't exist. Wow, the world just got a LOT smaller. What a load of B.S.!!
You said, " we don't have the time
nor space to review everything on the market. " Cmon, you're not some guy running a grass roots publication from his garage (I don't think). Many of your readers including me have some of the SSD's that have performance that falls between the two that you tested.
You said " this review was originally
published in the March 09 issue, which means it was reviewed around
early January. Print has a long lead-in time. " The large number of fast SSD's have been available much longer than Jan.You said "but time travel is sadly not yet
possible. " Now you jut seem like a mouthy little smartass Nathan!
The fact is you and the publication you work for are not doing a very good job. The information you provide is not valid, it's incomplete and by your own admission.....dated. Actually, even if this information had been published in Jan. rather than Mar. it would still be incorrect and dated.
And how would you know if anyone makes outragious claims about an SSD? You obviousley don't have the time, resources, or inclination to actually be able to compare what the market has to offer.
It seems that between myself and a couple friends, we posess, use and benchmark more current tech than an entire publication and it's staff.
You get a LAME-ASS Award MINIMUM PC!
I really believe that the shareholders of the publishing company of Maximum PC need to rethink this publication and make it what it once was. It used to be at the head of the class leading the way among all other computer enthusiast magazines. It had that "edge" and the most current up to date information contained within it's pages. Get back to your roots and advertising revenues will increase as well as readership. Like so many other business's today, a complete overhall would remove the stagnation and bring a new energy to Maximum PC.
They need to stop making
Submitted by Keith E. Whisman on Thu, 04/16/2009 - 1:30am
They need to stop making these low capacity SSD's and flood the market with high speed SLC High Capacity SSD's and I mean 500gigs and up.
If they can mass produce high capacity SLC SSD drives and actually sell them then prices will come down closer to the levels of regular hdd's. I want SSD's to replace HDD's but it's not going to happen until they make it into a why bother kinda of deal. What I mean is why bother using HDD's when SSD's are the same price and have the same capacity but are alot faster.
We aren't really going to see anything good come from this until the price of SLC Nand chips come down and that will only happen when demand gets high and more and more producers of these nand chips pop out of nowhere like everything else.
I agree...but
Submitted by Terabit on Sat, 05/09/2009 - 3:59am
While I agree with you that it would be nice to see a lot more high capacity drives for cheaper, but that won't happen for at least an year, since the cost of NAND memory needs to go down, and until they develop better manufacturing methods, that won't happen.
2X EVGA GeForce GTX 285 1GB (Each) | Intel Core i7 920 | OCZ Gold 6GB DDR3 1600 | Asus P6T Deluxe V2 Mobo | Seagate Barracuda 1TB HD | Corsair 850watt PSU | Antec Twelvehundred Black Full Tower Case | Asus VW246H Black 24" 2ms Monitor |
Um. Just because the
Submitted by TheMurph on Thu, 04/16/2009 - 10:51am
Um. Just because the market has 20, 500GB SSDs on the market doesn't mean that the raw costs of the NAND flash memory used to *make* said SSDs is going to get any cheaper. It's not like the manufacturers can just say "oh hey, we're going to all band together and flood the market with drives, which should magically reduce the cost of all the flash memory we need to buy to make these drives a reality."
Maximum Pc? I think not!
Submitted by rob41 on Wed, 04/15/2009 - 4:56pm
Good god! I've come to realize you guys at MPC really have no clue.
You said: "The CSX SSD’s read performance of 114MB/s is second only to that of Intel’s X25-M"
Before you guys start flapping your lips again why don't you actually get the basic facts.
There are at least a dozen models of SSD's (and several manufacturers) at Newegg alone that have better read speeds than this unit that you stated was second only to the intel X25.
If you were just the average blogger I'd cut you some slack but you don't deserve it. You're supposed to be the experts and inform people about facts.
I'm not even going to post here anymore. I continually see bad advise from the "doctor", horribly conducted reviews, and plenty of false information.
You get far from a "kick-ass" in fact you get a 1!!!
I wish I had a job like you Nathan Edwards, ya get it totally wrong and you're still gonna get paid!
To clarify, whenever we make
Submitted by nedwards on Thu, 04/16/2009 - 10:41am
To clarify, whenever we make product comparisons like "Second only to," we're referring to our testing. If we haven't tested something with our own benchmarks on our own rigs, it doesn't exist for us. Anyone can make outrageous claims about read and write speeds, and people get different results based on different testing protocols. And unfortunately we don't have the time nor space to review everything on the market.
More to the point, as Keith mentioned, this review was originally published in the March 09 issue, which means it was reviewed around early January. Print has a long lead-in time.
Since this drive was tested, we have seen several SSDs with much higher read and write speeds, but time travel is sadly not yet possible.
Man why are you hating so
Submitted by Keith E. Whisman on Thu, 04/16/2009 - 1:37am
Man why are you hating so much? He gave this drive a 4. That is pretty low.
If you think you can do better why don't you start a hardware review website and a computer magazine. BTW this article is really old. It was in the magazine and this article is really old because of the lag from the time the magazine gets written and printed to the time it gets to your door step. I heard someone say it's 3months. A January issue would be written in October or November for delivery in January. Alot happens everyday let alone that length of time and I don't think there are any magazines worth their salt that has a faster cycle.
I say Nathen you did a great job on this article and I'll stand by your score of 4 for this drive even though I want what I've been ranting about like above.
Feature
Review
Feature
Feature
Feature






