News - OCZ ramps up the IOPS

[HEXUS]

Debuts Vertex 3 Max IOPS edition SSDs.

Read more.

[Xtanto]

"75,000 IOPS (120GB) and 65,000 IOPS (240GB)"

This must be the wrong way around.

[semo]

You get these numbers now, but after the reviews are done OCZ will swap out the internals withouth changing the drive label/version.

I don't know how this dodgy company managed to recover from the V2 25nm transition fiasco.

[Mossy]

You get these numbers now, but after the reviews are done OCZ will swap out the internals withouth changing the drive label/version.

I don't know how this dodgy company managed to recover from the V2 25nm transition fiasco.

OCZ are hardly dodgy they were first to move over to 25nm and being the first they took the brunt others followed and learnt by their mistake, being the company they are by being first to market with leading edge technology might be the reason they can recover from a situation as such.

Moving to 25nm means in the future the process becomes cheaper and this price can then be passed on to users.

best knowing the facts before accusing a company of being dodgy.

[semo]

Of course they are dodgy. How can you change the internals to such a degree that you not only affect speed but CAPACITY as well and not change the labelling?

Affected customers have not been contacted. A mass recall has not been issued. There are people out there thinking that they have a drive they've read reviews about where in reality they are slower or in some cases SMALLER.

This is in no way a technological issue. This is an entirely a consumer issue. You can not label and sell a 60GB drive when in fact it is 51GB (it should be 56GB when formatted in Windows).

Corsair just put an A at the end of their model numbers to indicate the change of internals. Nobody is arguing that 25nm is the way forward despite its shortcomings. OCZ's dodgy dealings is an entirely consumer one. Imagine buying a 1kg bag of strawberries and finding that it is actually 800grams when weighing it at home. The manufacturer of that 1kg bag of strawberries will have absolutely no way of excusing themselves and will have to admit their mistake. OCZ hasn’t because they haven’t issued a recall of the affected products.

So in conclusion, OCZ ARE A DODGY COMPANY.

[badass]

OCZ are hardly dodgy they were first to move over to 25nm and being the first they took the brunt others followed and learnt by their mistake, being the company they are by being first to market with leading edge technology might be the reason they can recover from a situation as such.

Moving to 25nm means in the future the process becomes cheaper and this price can then be passed on to users.

best knowing the facts before accusing a company of being dodgy.

I don't think of OCZ as a dodgy company, however there is no excuse whatsoever for the 25nm fiasco. It shows extremely poor judgement by everyone involved in agreeing to perpetuate this lie.
Calling 115GB Drives 120GB drives is simply inexcusable. It is a lie, plain and simple. The pathetic excuse that they both have as much raw NAND (128GB) is plain wrong. Otherwise why are they not quoting the raw NAND capacity for their drives. Every other manufacturer uses the user available capacity for their capacity ratings. This includes HDD manufacturers where they also have spare area on their drives.

Another way of looking at it. By no measure on ever used on this planet, do their drives have 120GB of capacity. They have 128GB, 115GB, 112GiB or about 110GiB formatted.
How can they call the drive 120 GB?

[semo]

I obviously agree with most of what badass says but OCZ is definitely a dodgy company. The original OCZ was an atrocious company before they underwent a lot of changes about 10 years ago (back then they didn't retail in the UK market so didn't have a chance to build up their bad reputation).

Also, OCZ's RMA address in the Netherlands!
http://www.ocztechnology.com/support/customer-faqs/
Retailers in the UK are only obliged to supply warranties for the first year. So if your Vertex dies 2 years after purchase, you will have to send it to the Netherlands.
I see no reason why anyone should buy from OCZ. Their only unique products, for now, are their low cost PCI-e SSDs.

Since, there are so many options around, I see no reason why anyone should tolerate OCZ.

[Singh400]

^ Wow hate OCZ much? I for one would buy an OCZ SSD drive tomorrow if I had to. I trust them. Why do I trust them? They actively respond to feedback good OR bad.

[Perfectionist]

Corsair just put an A at the end of their model numbers to indicate the change of internals. Nobody is arguing that 25nm is the way forward despite its shortcomings. OCZ's dodgy dealings is an entirely consumer one. Imagine buying a 1kg bag of strawberries and finding that it is actually 800grams when weighing it at home. The manufacturer of that 1kg bag of strawberries will have absolutely no way of excusing themselves and will have to admit their mistake. OCZ hasn’t because they haven’t issued a recall of the affected products.

So in conclusion, OCZ ARE A DODGY COMPANY.

Intel do the same thing though. What company DOESN'T do it? It seems like practices on accurate advertisements of SSDs have yet to reach an agreed trading standard yet. Just made a post earlier about it, actually: http://forums.hexus.net/shopping-ret...ml#post2074087 - The tiny changes in model names were done all over the intel stuff while advertised with the same name, incredibly misleading for many people, but yeah.

Seems to be normal from what I've seen, it's a similar thing really to how SSDs (not just retailers but the actual producing companies as well) advertising sequential speed over random 4KB IOPs, even though 4k IOPs is a way more accurate measurement, the bigger sequential number is chosen because it sounds better. :/ Paul, I don't get why sites like Scan don't at least have the IOPs in the tech specs rather than completely hidden? Would help a lot of people

[badass]

Intel do the same thing though. What company DOESN'T do it? It seems like practices on accurate advertisements of SSDs have yet to reach an agreed trading standard yet. Just made a post earlier about it, actually: http://forums.hexus.net/shopping-ret...ml#post2074087 - The tiny changes in model names were done all over the intel stuff while advertised with the same name, incredibly misleading for many people, but yeah. Seems to be normal from what I've seen, same as advertising sequential speed over random 4KB IOPs, even though 4k IOPs is a way more accurate measurement, the bigger sequential number is chosen because it sounds better. :/ Paul, I don't get why sites like Scan don't at least have the IOPs in the tech specs rather than completely hidden? Would help a lot of people

There is a difference here. Ignoring relevant metrics and advertising largely irrelevant ones is not outright lying. It's like how car manufacturers used to love categorising their cars engines by their capacity, even though it only loosely affects performance. Virtually all companies do it. What OCZ are still doing is outright lying to their customers.
The performance difference is not as much of an issue in my mind TBH. The headline speeds are more or less the same

[Perfectionist]

Ah, edited my previous post to make it a bit clearer, was talking about both issues actually - advertised names of products being the same while having vastly different hardware (so people needing to rely on the small print model numbers to tell the real versions apart, which most people don't notice) ... What I'm saying is that Intel do it too, they all seem to do it. Is there any company that is actually upfront about the differences between models? I know there's the kind of renaming/rebranding issue with graphics cards too sometimes, it's worse for them becuase they don't seem to have easily distinguishable model numbers to know what's what, so if they change the name you basically just have to google?

... and ...
... the *other* issue of all companies seeming to advertise the sequential write speed while ignoring the stat that reviewers seem to universally agree on as being the most important (random 4kb Input/Outputs Per Second (IOPS) - p.s. I know you know what IOPs is, not being condescending there, that's for other readers' benefit, I've googled enough to try consider taking into account lurkers stumbling across threads) - I kinda see what you mean, like I said it's a bigger number, yeah, but companies should at least include the 4kb IOPs in the tech specs really if not in the main advertising, at least have the info available.

[semo]

Intel do the same thing though. What company DOESN'T do it? It seems like practices on accurate advertisements of SSDs have yet to reach an agreed trading standard yet. Just made a post earlier about it, actually: http://forums.hexus.net/shopping-ret...ml#post2074087 - The tiny changes in model names were done all over the intel stuff while advertised with the same name, incredibly misleading for many people, but yeah. Seems to be normal from what I've seen, same as advertising sequential speed over random 4KB IOPs, even though 4k IOPs is a way more accurate measurement, the bigger sequential number is chosen because it sounds better. :/ Paul, I don't get why sites like Scan don't at least have the IOPs in the tech specs rather than completely hidden? Would help a lot of people

Haven't you followed the story with OCZ? They have many versions of their V2 drives. The problem is, they are all called Vertex 2 and the customer has no way of knowing which is which. When they change the NAND chips it usually results in lower performance. It is measurable but it can be argued if it actually makes a big real world difference. When they first moved to 25nm NAND chips, not only did the performance go down but so did the CAPACITY. So let's say you bought a 60GB V2 drive in June 2010 and then another one in January 2011. The first drive will format to 56GB and the second would probably format to 51GB. Not only have you been lied to on the label but you can not RAID0 those two drives. If you found out by your own research then you could contact OCZ and try to get the situation rectified. If you didn't know anything about the fiasco, then well.... you wouldn't know about the fiasco and that you've been short-changed... and you wouldn't know about it... you get the idea. OCZ has not issued a recall. The issue has been buried and is no longer causing them financial loss.

Anand covered the issue but the general consensus around the net is that he went super easy on OCZ so that he doesn't burn bridges.
http://www.anandtech.com/show/4256/t...3-review-120gb

http://forums.anandtech.com/showthread.php?t=2143367

[semo]

Perfectionist, on the point you make about IOPS. The figure does not play a significant role in general usage patterns (office, internet, playing a/v, etc...).

I don't agree with a lot of people on this subject as they say "IOPS doesn't matter in everyday use", to which I answer "your everyday use is different to my everyday use". This is the reason why the Intel 510 is popular now (low IOPS compared to other SSDs but high sequential read speed).

Also, when OCZ screwed up the 25nm transition, the botched drives had a significantly lower IOPS than the advertised specs. Other companies attempt to bamboozle the consumer with an endless list of models and rebranded products. You couldn't buy two Intel Core 2 Duo E8400 CPUs which have different specs so you can't compare what Intel do to OCZ's fiasco