SSDs - State of the Product?

My SandStorm RAID 0 is acting up. Sometimes it is not recognized at boot and then some files are corrupted next time. It is fine for days if not restarted.
 
This thread's a bit long - which brand/model was that, Lunar?

I notice that TechReport and Anandtech have contrasting results with their SSD comparisons. TechReport thinks the Corsair Nova is great and Sandforce really sucks, while Anandtech thinks Nova is weak and Sandforce rocks. In particular, Sandforce bogs down badly on TechReport's single threaded copy tests.

What's the truth?
 
Thanks for that. I've looked at the websites for Corsair, G.Skill, MachXtreme, Patriot and Solidata. Only MachXtreme appears to be offering updated Sandforce firmware; it's 3.20. This makes me lean very heavily towards the firm that actually bothers to pay for production-ready updates and make them available to their customers. :-x
 
I don't know if that is good or not, but I don't want any firmware on an SSD.
 
The Corsairs are shipping with beta (or alpha really) firmware. To make matters worse, they're still not able to offer an update to fix exactly the sort of issues you described.
 
Well, nobody mentioned that a couple of months ago when I bought them. :confused:
 
I broke the array and will try one drive at a time for a few days. What else can I do? I'm not seeing anything new or better now compared to the same old StormForce, etc. SSDs.
 
So far the single SSD is not losing data. They are both connected, but one drive letter is removed. ;) I could try the other one for a few days. Would it help to secure erase the Cosair drives once in a while?
 
It is amazing how slow SSDs are increasing capacity and coming down in price. :cursin:

Why is it amazing? Flash memory manufacturing processes do not continually change. It is a stepwise progression. Max capacity is not really the issue. Several TB of layers will fit in a 2.5" drive if one is willing to pay.
 
Well that's a good point, where is the revolutionary stuff promised with these new Intel drives? The increased write life could be due to more intelligent firmware - that's what Sandforce is really about. And all manufacturers will have access to the new smaller-die flash.

On the face of it, Intel is promising to come out next year with something almost as good as Sandforce released a year earlier. It seems to me that no-one is interested in adding more channels because motherboards are still limited to 3Gb/s.
 
It is amazing how slow SSDs are increasing capacity and coming down in price. :cursin:

It's frustrating, but they are coming down. Price per gigabyte is now $2 at Newegg for 120GB Sandforce drives.

Spare a thought for people in countries such as New Zealand - they have to pay 30-40% above that.
 
Spare a thought for people in countries such as New Zealand - they have to pay 30-40% above that.

Newegg just kills you, doesn't it?

I think I've purchased more stuff from Amazon and Provantage this year than I have from Newegg. The flat $80 for two-day delivery on everything I order from Amazon saves me hundreds of dollars a year, easy.
 
Wow, it's rare that breaking news is so relevant. Nice pickup Bozo.

So the Sandforce 2000 series pretty much doubles performance across all parameters.
Anandtech said:
SandForce will have nearly maxed out 6Gbps SATA on its first attempt.

Says it all really. Makes a complete mockery of Intel's 'roadmap', reliability included.

On the other hand, there's this chilling prediction:
If the 3rd generation X25-M is really only capable of 270MB/s reads and 170MB/s writes, I’m not sure if there will be the motivation to deliver a 500/500MBps part into the enthusiast market.
 
Anandtech said:
At full speed you could copy 1GB of data from a SF-2000 drive to another SF-2000 drive in 2 seconds. If SandForce can actually deliver this sort of performance I will be blown away.

Me too.
 
Is it certain that the new generation of X25-E will only be 3Gbps?

"The SF-2600 ships with the external SAS bridge and a special firmware revision to enable support for non-512B sectors."

Looks like what you want for a SAS 6 RAID controller. Intel are still implementing it in future logic chipsets, aka way behind the 8 ball.
 
Let's hope the latest SandStorm is more reliable. It's not so fun to reformat every half-dozen boots. I'm not convinced that the transfer rates are realistic either, if they continue the multiply by two theory of compressible data.
 
Let's hope the latest SandStorm is more reliable. It's not so fun to reformat every half-dozen boots. I'm not convinced that the transfer rates are realistic either, if they continue the multiply by two theory of compressible data.

Yeah, the question is how it does with already compressed RAWs/NEFs/TIFFs etc.
 
Let's hope the latest SandStorm is more reliable. It's not so fun to reformat every half-dozen boots. I'm not convinced that the transfer rates are realistic either, if they continue the multiply by two theory of compressible data.

I must assume you have an outlier there Lunar. I have lots of SSDs doing stuff, some of it very demanding (Oracle databases, Citrix servers, VM Datastores, firewall cache, my gaming machine). So far no errors, no data loss.
 
I didn't mean that you had chosen an outlier, but that your drive (and it's poor reliaility) is not representative of the drives in general. Just bad luck.
 
That's really interesting Ddrueding. From your posts, I get the impression that you have mainly OCZ drives? Checking their site, I see that it's simple to download the latest (custom) firmware for their Sandforce drives (if needed).
 
I didn't mean that you had chosen an outlier, but that your drive (and it's poor reliaility) is not representative of the drives in general. Just bad luck.

From what I gather, he's reformatting to regain drive performance after heavy writes cause performance to hit the shitter?

The big downside I see to the SSD currently is that it is a variable speed device, that degrades with use. A finite life if you will.
 
Since a couple of weeks ago the array becomes corrupted frequently and needs to be reformatted. For the past week I've been running one drive only without issues. I'll try the other drive for the next week.
 
That's really interesting Ddrueding. From your posts, I get the impression that you have mainly OCZ drives? Checking their site, I see that it's simple to download the latest (custom) firmware for their Sandforce drives (if needed).

Indeed. I do have some X25-M, -M(G2), and -E drives, but most of my SSDs are OCZ. Vertex and their PCIe variants. I have no performance issues, so will not be updating the firmware.
 
Will the Samsungs even be available throught the usual channels, or only to the OEMs?
 
Since a couple of weeks ago the array becomes corrupted frequently and needs to be reformatted. For the past week I've been running one drive only without issues. I'll try the other drive for the next week.

No problems so far. I hope it is not a RAID issue.
 
Kingspec, POJ, 30 gig, update.

I have one that works fine. The other, I tried in a used, CF-51 Mark III Panasonic laptop.
Gives wonderful numbers. 70 MB/sec in ATTO and others.
Great Access times, HD tach, .02 MS.
Problem is, the drive sucks. It's slower with 2003 on it, in a Duo then a 7200 rpm seagate drive, and, that drive is NOT the top of the line. Huge lags, terrible performance, but, it looks ok on numbers.

Guess you pay your money, and, you get what you get...
 
Kingspec, POJ, 30 gig, update.

I have one that works fine. The other, I tried in a used, CF-51 Mark III Panasonic laptop.
Gives wonderful numbers. 70 MB/sec in ATTO and others.
Great Access times, HD tach, .02 MS.
Problem is, the drive sucks. It's slower with 2003 on it, in a Duo then a 7200 rpm seagate drive, and, that drive is NOT the top of the line. Huge lags, terrible performance, but, it looks ok on numbers.

Guess you pay your money, and, you get what you get...

I hope the amount was not much.
 
No, it is in TB. The X25-E drive is almost two years old. My X25-M G2 has 1.32TB, but it is used for the x64 OS/apps without any temp or page files.
 
Back
Top