Intel`s next SSD, 520 - Includes Sandforce

Guys, it is final. Intel 520 Cherryville SSD will feature a Sandforce controller. Apparantly it will be the same controllers as used by OCZ...

Recent news from Sandforce showed that they have fixed almost all the problems on their SSDs (Vertex 3 etc) with a brand new firmware. Could it be that Intel have been helping? Don`t know if it is true though

Anyhow, bad or good news?

Intel 520-Series Cherryville SSDs to Include SandForce Controller - Report - Softpedia

 

Good news if Intel and SF actually fixed the problems or Intel makes their own custom firmware and fixes things for them but leaves OCZ with their problems .

 

muhahaha yeah that would have been fun. "Hey guys, look at our drives. No problems." while OCZ BSODing all over the place

 

...screw that. So much for Intel's famed reliability if they start using SF controllers.

 

not sure if that linked article says definitively that the 520 uses an SF controller. a clevo reseller (mythlogic) said in a post that he had definite knowledge (covered under a NDA) that it did not use an SF controller.

 

It all points towards a SandForce SSD. The rumours long ago. The typical SF capacity, 120GB and 240GB. The same 4k speed as all other SF drives get. The same read and write. And now this.

I would like to see the post from that reseller

 

Totally agree!

 

I still had hopes for new Intel controllers.... but sandforce.... ill stay with Crucial M4 and probably will continue with Crucial from now on.

 

this is gonna be interseting.....

 

Cloudfire, this is far from 'final'.

That doesn't look like a news report as much as it looks like it is simply repeating Swedish gossip.

Even if Intel uses SF controllers - I still believe very much in the power of the Firmware as much as the philosophy of the company providing the SSD drive for a specific market. Not that I will blindly recommend them - just that they will be given a very serious look if and when they ship.

Still, with all the 'non-information' we have, I still doubt that Intel needs SF to give us a new SSD product at this time.

Their Marvell firmware/controller combo is still the best overall speed/reliability champion - why would they go 'begging' to SF for their weaker and much less bulletproof 'last year' design?

Now, this could be the third gen SF controller that we could see debuting, but I would have thought we would have heard something about that (by now).

 

Where is that reliability? 320 series still suffer from the "bad context blah blah" firmware error and it hasn't been fixed yet.

Intel sent out an updated firmware which made things even worse... You will brick your hard drive if you run full diagnostic scan of Intel's SSD Toolbox with the so-called updated firmware.


--

 

Until there's news detailing how SF has replaced & upgraded engineering staff, I'd tend to view this as 'OCZ, part II, Intel version' and probably stay away, at least until these SSDs have a proven track record.

Bitten by OCZ once, not gonna happen a second time.

Fortunately, for my needs a 510 would be just fine, so I'm in no hurry to jump onto the bleeding edge bandwagon here.

 

Sweclockers, which is the website that is quoted, isn`t just a vague gossip site. It is like Anandtech, hothardware etc and have given us exact information about upcoming products before. Like the US price of the 320 long before it was released. And it is not gossip, they have actually seen a intel document where the 520 was described in detail with a Sandforce controller mentioned.

And yes, I do believe Intel have something to gain by jumping on to the SF wagon. Truth is, that Sandforce drives such as Vertex would have been awesome if it weren`t for that reliability problem. Speed wise, in terms of benchmarks etc, they are #1. No doubt about it. And non technical people are buying in to the SF marketing, because they don`t know that Intel 510 is equally fast although it lags behind in benchmarks to Vertex 3. All the reviewers have been praising Vertex 3, and you have to really look between the lines to actually see that 510 and M4 is equally if not better in real life scenarios. Like compressed/uncompressed data, power consumption, OCZ history and customer support, etc etc. Heck, even people with a bit of hardware knowledge doesn`t care about this, hence why SF drives sell like hot potatoes.

Now if Intel can use their advantage as a motherboard/chipset producer and their experience with Marvel and Intel controllers, to make a firmware their way, combine it with the latest reports that SandForce have almost ironed away most of the SSD problems, I think we may have a winner here. Perhaps that is what Intel saw before making the decision.

 

Look at This and This and tell me that 320 suffers with a reliability problem. The problems you are quoting are not wide spread and a very small percentage are actually having problems. Google or look at newegg for any SF product you like, compare, and tell us again that Intel have a reliability problem...

The exception is Kingston HyperX SSDs that also have a SandForce controller. They have an astonishing history where it looks like they have made a very fast and reliable drive. Which is the reason why I am "defending" Sandforce a bit on the post before this one. Maybe they have ironed out many of the issues indeed

Here is the Newegg ratings of HyperX SSD
Newegg.com - Kingston HyperX SH100S3B/120G 2.5" 120GB SATA III MLC Internal Solid State Drive (SSD) (HyperX Upgrade Kit)

 

Checking the reviews, there are some that reflect BSODs also,

But only 2/34.

If crucial m4 didnt existed, maybe i would have tried the intel+sandforce, but ill let others test and see if intel makes it right.

 

I guess the only drive to get is made from Samsung or Micron.

 

I don't claim to know what site was quoted in your link, but with lines like this:

Which should obviously read as 'If this report is indeed true...', I stick by my assessment that this is simply repeating idle gossip.

I put that article in the same brain-space as this:

See:
http://forum.notebookreview.com/8010912-post5.html


Why don't you support that hypothisis?


Either one of them (equally) could turn out to be true, but they are both things of the 'I heard he said' nature at this point in time.

The link you provided simply wants more hits to their website - doesn't matter if the 'news' reported is true or not - any attention is good attention (in a marketer's world).

 

So much suspense, I can't wait to hear what new on Intel's upcoming SSD's in the coming weeks. Rumor is Cherryville is slated for a early November release, personally I'm waiting on Hawley Creek and hoping for a mid of November release on the new line of mSATA's.

 

So whats the whole hoopla about sandforce controllers? Is it because OCZ uses them?

EDIT: Reading up on marvell vs sandforce.....

 

that's awsome...
for the first time i am seeing such good reviews for 3rd gen SF product....
looks like kingston got it all figured out finally..............

 

3rd gen SF? Where do you see that? (You don't think that the HyperX is 3rd gen, do you?).

 

Looks like it but I don't want to put words in his mouth. The Kingston is fairly recent though isn't it? I mean I know its been out for a few months as of this moment but compared to the Crucial M4 and the OCZ SSDs I'd say its still sorta "newish".

Anyways I wish Intel would release this like now so one could compare and contrast with the Samsung PM830 line. I'm guessing that Samsung uses the marvell controller?

 

^^^ from what i understand, samsung uses their own controller (intel apparently used to do this before their latest effort), which is supposed to be one of the attractions, namely, that the 830 is an all-samsung product, and doesn't rely on any outside manufacturer, and so quality control should be easier.

 

I've never had any sort of problems with my 320... Even when I run said scan on it, mine still works.

 

Even better. Thanks for the info.

 

Well it is -kinda- new, meaning the HyperX SSD have been out there for almost 4 months now. Vertex 3 got bad reviews from the very start, few days after it was listed at Newegg so the stats you see from HyperX is valid imo.

Samsung is one of the few that have their own controller like mentioned above.


Here are the news from Anandtech about the Sandforce fix recently pushed out. I don`t know if it truly took care of the problems though:
Anandtech - Sandforce Indentifies Firmware Bug Causing BSOD Issue, Fix Available Today

 

I`ve read the source, the article from Sweclockers. It is written in swedish, a language I understand. Anyhow, you are right that it is not final per se. It is only final when Intel say it themselves that 520 will feature a Intel controller or SF whatever. But I choose to believe, or make up a high percentage probability, that the next SSD from them, will be Sandforce. Could be dead wrong, could be right. Only time will tell. And it is not very far away. Last news I heard was that Cherryville is supposed to release in November 11th

 

Even though I really don't like Samsung, I'm tempted to get their PM830 series drive, since they do make every part of the drive... (heck, I just found out the Samsung group makes seagoing ships, too!!).

 

Why You don't like Samsung?

 

don't study politics? ever.

 

enlighten me.....

 

Okay, I'll enlighten you:


Kingston HyperX main specs:

SF 2281 controller (same as V3 from how long ago...).

MLC nand (same as products from the last 4 years...).

Advertised performance specs within margin of error of all other SF based SSD's.

Kingston does not make it's own SSD's - it simply uses off the shelf available parts (including firmware, nand and other necessary parts...) and then puts it's label on it to make it 'theirs'.

SandForce controllers are not at 3rd gen yet and in my opinion barely reached 2nd gen status with the BSOD snafu from the last year or so. Hopefully the newly released firmware will fix these old controllers - but if and when SF 3rd gen does appear on the market, I hope people are more wary of them than they were with the current one.

See:
AnandTech - The SandForce Roundup: Corsair, Kingston, Patriot, OCZ, OWC & MemoRight SSDs Compared

From the above article (last page, final words):

 

And those 2 quotes say A LOT. Anand has long been a very clear supporter of Sandforce products, but even he is starting to back away from them.

 

Back away from them? Anand?

I feel as though he is putting the transmission into 1st gear at the top of the hill overlooking a lake as night is falling and OCZ and SF are passed out with blood trickling down from their heads and mouths - as he jumps clear and watches them disappear into the lake far, far below.

But, don't worry - he won't be convicted - his new friend will vouch for him (you know, that friend that drove out to those back roads to pick him up?) - they'll have an airtight alibi that they were together all night - at the celebration gala of Intel's new SSD introduction.

$%@$%!!! I guess I gave away who the 'friend' was!

 

I get it, I'm understating things

The point stands. If even one of the most vocal supporters of OCZ is now advocating Intel or ANYTHING that isn't Sandforce, that says it all.

 

Not sure it has been mentioned any where, SF has been bought by LSI for 322M. It is quite some money but can also mean they have problem for the original rumoured IPO plan.

 

Nah, SF's directors knew they had run SF into the ground - LSI is simply a way for SF to continue to live - at least in name.

I have also read that the controller will still be available 'widely' but if consumers are finally getting the message that SF is lies, damn lies, marketing (DuraCrap technology...) and reliability that is less than those finicky, flimsy mechanical drives that rely on actual spinning platters to save our data to, then that side of the SF legacy should quickly dry up in the retail channels, unless they have sales like HP did with their Touchpad 'fire sale' - but at least those people knew that the 'pocket change' product they were buying had no future.

See:
LSI Deal Won't Change Availability of SandForce SSDs

 

The only reason Intel would do that is if they had no alternative controller which leaves serious implications about their Marvell controller, did something go wrong?

 

Do what?

 

With their controller, the retail market is something they cannot survive. You can treat some reviewers well and have them praise your stuff but there is only that much s* out there.

the server space may be better, not that the server space required less reliability but that you have a much controlled segment(only certify and test certain MB/BIOS combination) which they have a better chance to survive.

 

Use a Sandforce controller in their SSDs...someone missed this thread, lol

 

lol...

Okay, so I dozed off for a blink-of-an-eye.

 

Use the sandforce as opposed to the Marvell controller. I don't know, if Intel does this I think it would have to be more of a forced acceptance than something based on their choice.

That's what I meant by it though.

 

Okay, thinking back on things (SF 'fixed' firmware, LSI buying SF and rumours that Intel will use SF controllers...)...

With LSI testing, validation and technology inside SF firmware and/or their controllers, maybe Intel is really comfortable enough with this new SF to try an Intel branded SF based SSD.

This could be the performance breakthrough (in storage sub-systems) that I've been dreaming about since I've seen what a single Intel 510 250GB SSD can do.

With the collaboration of the SF engineers (or, at least their blueprinted product), the LSI enterprise validation (reliability above all else!) clout and the keen way the Intel minds know what really makes a system tick (4K R R/W's are not all they're cracked up to be...), we could have another game changer on our hands.

Now, I'm really looking forward to this possible new product.

Even if it doesn't come now, it seems like it is destined to come at some point in the near future.

Maybe the LSI team really 'fixed' the firmware properly with the SF headaches? Just to be sure it could be done?

Maybe LSI forced the SF engineers to actually debug their own controller - before they put in their offer to buy SF?


Maybe, we can have an LSI mSATA option soon with built in RAID that is reliable, bootable and PCIe FAST?

Man! I want to see the next gen Ultrabooks already (not the mild Zenbook...).


Does anybody else see these things? Or, should I stop having my 1/2 milk and 1/2 vodka breakfast drinks?

 

Perhaps you should take it easy on the White Russians in the morning, at least wait until noon, geez...lol

But I'm as enthusiastic about this as you are, I've actually been hoping we would see a mSATA option that is at the very least backwards compatible to the current line-up of laptops. Perhaps these guys working together is what it will take for their to be such a product, however, I hope there aren't any type of backlashes.

I wonder how the SF controllers become so buggy, do this just end up making it and are like, it let's get out there because they want to be the first? I don't know, reliability is an important aspect of buying stuff, I'm not really much of an early adopter of anything really, even with new laptops, I'll wait 3-4 months after it is released in order to get it.

 

Hell just give me LSI Raid card that can send trim command to the raid array (SSDs) under windows 7 and ill be a happy camper.

 

White Russians! That's what they're called, I, uh, forgot for a second.

It seems like Intel may already have this piece of the puzzle solved for us with IRST v11.

See:
http://forum.notebookreview.com/solid-state-drives-ssds-flash-storage/619685-ssd-provisioning-3.html


Look at posts starting at #24 in the link above.

 

Since when did LSI fix the controller problems Sandforce had? LSI bought SandForce AFTER a firmware that took care of the problems were rolled out.

If anything, it is more likely that Intel paid a big role in this fix. Intel 510 never had problems with the sleep mode of SATA drives that OCZ had that resulted in BSODs once you woke the device/link up. Intel knew exactly how to write a firmware for it, and no wonder, since Intel make the chipsets.

I think LSI chose to buy Sandforce because A) It is well know for it`s speeds B) Sandforce themselves have been pushing out firmware after firmware that made more problems than it fixed. People were telling others to stay away from any Sandforce products. The word spreads around throughout the internet. This resulted in the stock price/value for SF falling down. I think the price that LSI paid was extremely low imo. The market never had a chance to pay attention to this which eventually would make the SF stocks/value climb up again. LSI have been wanting to buy SF for a long time, and they made their move before stocks/value went up. Nice deal indeed.

My guess

 

Cloudfire,

I was guessing at all this too (of course).

But just because LSI's acquisition was announced just a few weeks after the SF 'fixed' firmware was released, doesn't rule out the possibility that they were involved in that effort.

Don't worry, I don't think they did it for charity (if they did it at all) - I'm sure SF paid dearly to get professionals involved.

The stock was privately held so no funny business from that angle.

If Intel did contribute to this SF firmware fix because of inside knowledge of their chipsets, I am sure that SF would be pointing to Intel right now and saying 'See! I told you, it wasn't our fault'.

 

Nope that would be laughable and totally wrong if they did. You don`t make chipsets and motherboards for storage devices, you make storage devices for chipsets and motherboards.

The Sandforce problems are not hardware related, they are caused by software. There is a reason why OCZ was able to fix their Vertex 3 with a firmware...

Intel have an extended knowledge of the SATA links and the sleep modes, more than any other SSD producer. Hence why I think they got the whole LPM right, while drives like M4/Vertex 3 had problems. Users had to disable lpm to never let the SSD idle because that would cause a link to the SSD to power off/idle which again caused the BSODs and freeze remember? A later firmware fixed that as well.