Intel's 34nm NAND SSDs launch in two weeks

[One_Hertz]

IN THE REAL WORLD, where these particular SSDs should get used everyday, you will find there is little to NO difference between them.

That depends on your CPU/Ram speed. If it's like 4ghz then I am sure there will be very minimal differences. As you get closer to 5ghz, CPU bottlenecks get smaller and storage speed matters more.

[cx-ray]

I use both Vertexes and X25-Ms every single day, in multiple configurations at home and at work.

What about the maintenance work of the firmware? I've used an X25-M 80GB G1 as an OS drive for months without permanent performance degradation.

Also, reading about a possible 4 hour long firmware update doesn't sound very encouraging.

http://www.ocztechnologyforum.com/fo...ad.php?t=56948

[pokerezzet]

My friend went to Thailand on vacation so i borrowed his intel drive to do some tests with 3*Intel X25-M G2 80GB in Raid 0 on my Motherboards ICH10R Controller



Is one of the disks bad or is it the ICH10R controller limiting my read speed to stay below 700MB? What is the best Raid 0 stripe size to choose for best overall OS performance with SSD 3 drives?

[One_Hertz]

Why is his CM so pretty? I wish my CM was that pretty

Yes, ICH limits you to sub 700MB/s. Stripe size 128k is best.

[Brahmzy]

What about the maintenance work of the firmware? I've used an X25-M 80GB G1 as an OS drive for months without permanent performance degradation.

Also, reading about a possible 4 hour long firmware update doesn't sound very encouraging.

http://www.ocztechnologyforum.com/fo...ad.php?t=56948

Huh? The X25-M G1's DID need a F/W update.

I don't know anything about a 4hr F/W update. My F/W updates have taken less than 1 second to flash on the Vertexes. If it takes more time than that, you're doing it wrong.

I'm not trying to defend OCZ or INTEL here - I'm using both products. I am not brand-biased at all, I buy the hottest product out, whenever it's out. But yet, you're still nitpicking OCZ from the sounds of it, making up 4hr firmware updates. The thing takes literally 5-10 including all reboots. How long did the intel take?

[cx-ray]

I'm talking about performance degradation through general use. After firmware 8820 for X25-M a few months ago, it was basically an install and forget SSD. No need to run any wiper tool that could potentially corrupt your data or clean erasing. The drive kept itself in good condition.

As for the long firmware update, the info is on the OCZ website link I posted previously.

The point is the X25-M has been largely trouble free, while OCZ users had to go through quite an ordeal of "betaware" to get to where they're today. That is what contributes to my lack of confidence in their product.

[Rhys]

I'm talking about performance degradation through general use. After firmware 8820 for X25-M a few months ago, it was basically an install and forget SSD. No need to run any wiper tool that could potentially corrupt your data or clean erasing. The drive kept itself in good condition.

As for the long firmware update, the info is on the OCZ website link I posted previously.

The point is the X25-M has been largely trouble free, while OCZ users had to go through quite an ordeal of "betaware" to get to where they're today. That is what contributes to my lack of confidence in their product.

It doesn't always keep itself in good condition. It depends entirely on the workload, and that workload can still be described as general.

Shame that G1 ain't getting TRIM! because then there would be 0 issues with degradation.

Currently got my vertex's back in my desktop in raid 0 as the G2 delays are getting beyond the joke, as was the Indilinx launch.
Truth is I forgot how good a few of these puppies are in raid 0 (write back cache off). I actually think the 'average ssd user' MAY get more noticeable daily task performance out of a couple of 120gb turbos than the G2's!

[Rhys]

I just followed the instructions in HDDERASE 3.3 and used option 1 for a secure erase. After that I setup RAID 0. Then, before installing Vista I created a 128KB starting off set (Diskpart - create partition primary align=128). Once the Vista install was done I immediately installed the Intel chipset drivers, rebooted, shut down system and power. Next, the Intel Matrix driver, reboot, shutdown and power off.

Finally, I installed the video driver so I could properly test the sleep feature.

I don't know whether shutting down had any effect. I just did it out of precaution to ensure myself that all settings were saved. I also followed the same steps with the offset because it worked last time.

Is it still ok?
Little worried about mine now!

Alanis Morissette springs to mind from all the comments back in march.
Gunna join the G2 beta testers party this time Griff? Oh the irony Lol!
<- Looks like this banana's got stage fright!

Edit: oh remember Griff 'Quality' SSD's shouldn't be anymore risky than buying HDD's!
I can't understand why you haven't been jumping up and down on this issue already!!!

[Griff805]

Again Rhys... Intel has not stated that they will not release TRIM support for the G1 drives. So, get over it- You sound like a 3 year old who overreacts and exaggerates everything. CX-RAY and Earzz are having a similar issue but it is not anywhere near confirmed to be the drives themselves yet. The G2 has a single issue that 99% of all users would never have even noticed. On top of that, Intel has already addressed the problem. So, tell me how buying a G2 is a risk- go ahead, tell me.

Also, the G2 has only been delayed for less than 2 weeks- that is nowhere near the months that Indlinix was delayed. Intel has already released a shipment so now it's just a matter of supply and demand- the first batch was already bought out. That's different than a product launch delay.

Rhys, do you work in the media? You'd do a good job there blowing things way out of proportion and making crazy claims about things without any kind of information to back it up-

[ChaosAD]

I actually think the 'average ssd user' MAY get more noticeable daily task performance out of a couple of 120gb turbos than the G2's!

Can you explain this a little better please. I would like to raid 2xSSD's after summer vacation and i was thinking about the G2's.

[cx-ray]

At the moment things appear to be functioning properly again. I've put the pc to sleep 7 times now. One of the sleep sessions was a real one of several hours as opposed to the others which were only quick tests. I've also set the pc not to reboot after a BSOD so I can see what the error code is if it happens again.

[Ao1]

Have intel released a g2 firmware update for the drives that slipped out before the bios password issue became evident?

cx-ray, what changed to stop the issues that you reported? I would be surprised if the offset was the problem. This only occurred in sleep mode? Did you submit a support ticket to intel?

[Earzz]

Well i had the same problems as cx-ray, i fixed my problems with removing my ssd's and leave them out for a few minutes after i used hdderase.
The problem was not only in sleep mode, it also removed the RAID after the PC was turned off.
Now im running my G2's in RAID without problems anymore, so i keep my fingers crossed.
Thanks for the help guys.

[cx-ray]

cx-ray, what changed to stop the issues that you reported? I would be surprised if the offset was the problem. This only occurred in sleep mode? Did you submit a support ticket to intel?

I've only had the BSOD and complete loss of the RAID 0 volume once. That was after I used HDDERASE 3.3 and let Vista set the default offset.

The crash happened right after waking up from sleep mode for the first time. After that I used HDDERASE again and installed my stuff with the 128KB offset I initially used. Nothing else is different.

I don't know whether the offset caused it and didn't feel like installing Vista all over again just to find out Things appeared to work fine previously so that was one variable I could rule out. I now also have full and current back ups in case it happens again.

I think it's too early to submit a ticket to Intel. In my case it was only a single occurrence after all.

[Rhys]

Can you explain this a little better please. I would like to raid 2xSSD's after summer vacation and i was thinking about the G2's.

Similar to what Brahmzy is saying. From using both a vertex and x25-m there is zero perceived OS loading between the two. The only place you can really tell the difference is when you are creating/copying/saving large files such as iso's PSD files .RAR's etc.

When SSD's were too expensive to fit 'many' working files sequential write speed wasn't much of an issue. But now SSD's are getting bigger and cheaper many of us can now begin to work directly with files off our SSD's because they are now becoming big enough to do so, at least at a more reasonable cost.

I think people are now underestimating the importance of sequential speed, and the all round performance of a drive as a whole.

A couple of 120gb Turbos 'should' give you roughly the following.
240gb of space
550mb/s Reads
400mb/s Writes

A couple of 80gb G2's 'should' give you roughly the following.
150gb of space
550mb/s Reads
160mb/s Writes

Of course I'm not currently sure of the price differences there will be when both are 'fully' released.
Apparently the Turbos will be replacing all the Vertex's but the Vertex will still receive full support.

[Rhys]

Again Rhys... Intel has not stated that they will not release TRIM support for the G1 drives.

Griff come on!
All the reviewers that I'm sure work closely with Intel say Intel will not support TRIM on G1. I'm sure ALL of them didn't say this for no reason. And with the silence of Intel I think we can safely say they are not going to.

Maybe in the future they might well decide to support G1, i.e. if dumping the G1 is getting too much bad press. But apart from that chance it's obvious it's a done deal. A fool should be able to see the writing on the wall.

As for me working in the media, Lol!
The media is a propaganda machine designed to form your opinions for you, so you don't have to, the sheeple believe whatever lies their fed to just give reason/justify various actions or policies.

Yet I bet you still think they simply exaggerate stuff.

[One_Hertz]

I think people are now underestimating the importance of sequential speed, and the all round performance of a drive as a whole.

A couple of 120gb Turbos 'should' give you roughly the following.
240gb of space
550mb/s Reads
400mb/s Writes

A couple of 80gb G2's 'should' give you roughly the following.
150gb of space
550mb/s Reads
160mb/s Writes

Of course I'm not currently sure of the price differences there will be when both are 'fully' released.
Apparently the Turbos will be replacing all the Vertex's but the Vertex will still receive full support.

Few apps take full advantage of sequential speed, and the ones that do generally only take advantage of the reads...

[Rhys]

Few apps take full advantage of sequential speed, and the ones that do generally only take advantage of the reads...

Yeh I know, But for instance I spend most of may time working with large PSD files. Even though fast sequential speed may not be taken advantage of when loading apps/etc they are when your loading/Saving 2gb PSD files.

When I'm working with images ect, I like to hit the save button periodically encase my system crashes, When the files get to >2gb it takes ages to save and slows down your work flow, and generally interupts your train oh thought.
Big write speed makes your work allot more efficient.

Edit: Loaded a 1.8gb PSD in Photoshop CS4 (Photoshop was already loaded)
Drives
7200 rpm 1TB hitachi 75/ishmb/s writes in Crystal diskmark
2x Vertex 30gb Write back cache off 200ish mb/s writes Crystal

Load time
SSD 5.6 secs
HDD 29.9 secs

Save time
SSD 9.4 secs
HDD 20.8

[Brahmzy]

I'm talking about performance degradation through general use. After firmware 8820 for X25-M a few months ago, it was basically an install and forget SSD. No need to run any wiper tool that could potentially corrupt your data or clean erasing. The drive kept itself in good condition.

As for the long firmware update, the info is on the OCZ website link I posted previously.

The point is the X25-M has been largely trouble free, while OCZ users had to go through quite an ordeal of "betaware" to get to where they're today. That is what contributes to my lack of confidence in their product.

The situation on the OCZ link is far from common. 95% of FW flashers have no issues at all. It's cake.

The Vertexes have GC now which keeps them all clean. I think OCZ's philosophy differs a bit from intel. They are specifically aimed at enthusiasts, and they'll be the first to admit it - they have. They have provided almost monthly FW updates to their customers and have requested feedback. This has fine-tuned their product to be an excellent one, in many ways. This has basically used the consumer as a beta-tester, if you want to call it that. Their product came out solid and is even better today, in fact, the Vertex may stay cleaner than the intel SSDs do (at least the G1's, from personal experience) I have a RAID0 Vertex array that performs the same now as it did when I first installed it, better actually. And this is with heavy use and a lot of data on the array. It maintains it's write performance, thanks to GC.

So, intel's philosophy is less-frequent FW updates with more changes, OCZ's is more-frequent FW updates with fewer changes - that's how I see it. Intel has a much bigger in-house testing/beta mechanism than OCZ does. HOWEVER, OCZ and the other Indilinx vendors have direct input/communication for future FW fixes, where as that line doesn't really exist with intel with it's cutomer base, although it has not been needed as much.

We all know this is newer tech and these problems will not even be a concern a year from now as things mature. Both excellent products, but I have to give OCZ and Indilinx respect and credit for their hard work on taking steps to pefect this cutting-edge tech.

[Ao1]

Ocz are a niche company selling low volume cutting edge products at a high profit margin i.e. the enthusiast market. Nothing wrong with that but they obviously have to be very focused on customer support to retain that market. “Enthusiast” support i.e. how do I tweak this product, what’s coming next etc is something ocz excel in. Product support, outside of RMA, i.e. firmware updates, however comes down to who is supplying the product they are rebranding. With Indilinx it has worked out well for them and whist they can’t make changes to the product themselves they have been able to influence changes.

Intel on the other hand sell by volume and “enthusiast” support is non existent. On one hand that is disappointing for a new product but on the other hand Intel did a lot of work to understand usage patterns etc before they started design, which enabled them to release an optimised product to the market from day one. What they have been hopeless at is interacting with customers and explaining how best to use that product, but they are just not set up to do that. That’s not to say you don’t get support, they just do it via a support ticket rather than a forum.

[Rhys]

Ocz are a niche company selling low volume cutting edge products at a high profit margin i.e. the enthusiast market. Nothing wrong with that but they obviously have to be very focused on customer support to retain that market. “Enthusiast” support i.e. how do I tweak this product, what’s coming next etc is something ocz excel in. Product support, outside of RMA, i.e. firmware updates, however comes down to who is supplying the product they are rebranding. With Indilinx it has worked out well for them and whist they can’t make changes to the product themselves they have been able to influence changes.

I actually think ocz took a bit of a risk with Indilinx, as at the time it a new company made from allot of ex Samsung engineers and pretty much unknown to most of us.
I think as Indilinx was so new and hadn't had chance to build up any find of reputation yet, OCZ wanted more control, and Indilinx obliged.

It wasn't until the bugs were worked out that other manufacturers jumped on the bandwagon.

Luckily it all seems to have paid off!

I'm actually really looking forward to the next gen controller they will release because a) they should have learn't allot b) they should have plenty of cash now to spend on R&D.

[Ao1]

I've looked at Intels' site and there is no sign of a fw update for G2, which you thought they would have posted as soon as it was avilable for those that already have the drives with the password bug. It makes me wonder if returned drives are being flashed with new fw yet.

34nm is a big jump. It's not V2 it’s G2. Maybe around the same performance, but it's a leap in technology. I'm hoping that Earzz & cx-ray's problems aren’t an indication of a problem as a result of that. Until TRIM is available I don’t see any real advantage of the G2 drives (apart from price) so I don’t see why there would have been a big rush to get these drives out early. In other words if there are bugs it will be disappointing.

[JohnZS]

I think the firmware update might be available next week, I'm not sure on this....just something I read somewhere suggested so.
Have a read for yourself guys
Might even be out tomorrow?
John

[modboxx]

A general question on trim? will there be trim for vertex and intel drives in raid0?

I know vertexs got a wiper now but only for single drives.

[lowfat]

A general question on trim? will there be trim for vertex and intel drives in raid0?

I know vertexs got a wiper now but only for single drives.

depends on if RAID controllers have drivers/firmware written for TRIM.