Looking at recent prices I am amazed at the number of SSDs that hover around $100 CA for 128 GB after rebate. Mind you, the ones that I'm thinking about are from ADATA, who seem to have been making a big push recently, but other drives are starting to mimick their pricing.
OCZ still has a lot to prove to me before I'd ever touch one of their products again. Too long a history of buggy performance and crappy, blame-the-customer-first support.
I specifically boycott products that are only competitively priced through use of a rebate. I understand the financial side of using a rebate vs a sale price from the vendor's perspective, but until they make great strides in making the process less hostile towards the customer (forms "lost", rejected over insignificant errors or imaginary reasons, 8 week processing times etc) they are an absolute deal-breaker for me.
Give me an automated system that accepts an online form, optically recognizes a scanned image of the product UPC code, and accepts a scanned image of the receipt to be audited at random, which INSTANTLY transfers funds to you through paypal or interac email money transfer and we will talk. <end rant>
Looks like the SSD Bench section is missing quite a few other current drives besides just the Vertex 4's: - Samsung 830 (the "best" mlc) - Crucial m4 128G (one of the most popular @ $125, though slower on writes) - Intel 520 - Kingston HyperX - Mushkin chronos - Sandisk
240GB Kingston HyperX and HyperX 3K are in the bench.
We don't have any SSDs from Mushkin yet, hence no data in the bench. However, due to the popularity of Mushkin, I'll try to get my hands on their Chronos series ASAP, but it will still take awhile.
We'll have a 480GB SanDisk drive in the house in a week or two. However, keep in mind that all SF-2281 based SSDs are pretty much the same, thus we don't test every SF-2281 based SSDs that's out there.
Also, keep in mind that some models may be missing some tests, i.e. we only ran the most essential tests and left some out to save time.
Ah - yes - thanks for clearing up why I thought some drives were missing. The default 'Anandtech Storage Bench' just had the fewest drives, and I lazily ctrl-f searched the page ala cpu passmarks searchable list.
I agree. Right now it defaults to 2010 Light Workload - Average IOPS test, which is something we don't even use anymore. I sent Anand a message about this and will let you know what's his take on this.
What? That happens for EVERY computer component! Have you seen how awful today's low-price laptops are? And rotational hard drives, printers, power supplies, ....
It's not just computer components of course, this applies to everything sold at Walmart and most other stores for that matter. Crap quality for cheap price.
I don't really get what you mean. If you are referring to the fact that the 64GB drive uses 8 NAND packages instead of 16, that's the case with all 64GB SSDs with 2Xnm NAND. 2Xnm MLC NAND die is 8GB, so that limits the maximum amount of NAND packages to 8 in a 64GB drive.
OCZ has a release candidate for firmware revision 1.4. Apparently, it gives a serious boost to write speeds. I'd really like to see these drives retested and benchmarks for 64 GB version with this firmware.
We'll most likely wait for the release version before publishing any data. I talked with Anand earlier today and he said he has done tests with the 1.4 RC but there are still a few things that should be fixed in the final version, if I interpreted his messages right.
A French site has been tracking faulty returns to a big French e-tailers for a few years. SSDs in general are not very reliable, and OCZ SSDs in particuliar are crap: http://www.hardware.fr/articles/862-7/ssd.html
The fact that OCZ is that much worse than Corsair which is selling the IDENTICAL PRODUCT (Sandforce controller plus some NAND on a PCB) is really damning. Just awful. Yuck!
While I'm not saying that the data is totally irrelevant, I would like to point out a couple of points.
1. OCZ Vertex 2 is experiencing the highest return rate but it's no longer a high performing SSD in today's standards. The return rate of Vertex 3 is noticeably lower, although still fairly high.
2. OCZ was the first one to use SandForce's second generation (SF-2000 series) controllers and they had exclusivity for months. Some of the returns are due to early firmware flaws and bugs, which were fixed before others started selling SF-2000 SSDs.
3. There is no breakdown of Corsair's faulty SSDs. One should note that Corsair sells Marvell based SSDs as well, so their SandForce based SSDs could have a high failure rate but their Marvell SSDs would balance the return rate.
4. There is no info on return rate of Indilinx Everest based SSDs, which is what Vertex 4 is. Now OCZ has more freedom when it comes to firmware and they are not at the mercy of SandForce. OCZ has also said that they are spending more time and money on validation now.
5. It includes only four brands so we don't really get a good idea if OCZ's return rate is due to SandForce or is the problem at OCZ's end.
Again, just pointing these things out. However, I don't think that data is completely valid when talking about Vertex 4.
How dare you use logic and reason against someone's righteous anti-OCZ crusade??? :)
Seriously, I would tend to agree. I remember when the original Vertex and Agility drives came out, people loved them. They went downhill when they started selling SandForce drives, but Vertex4, Petrol, and Octane are non-Sandforce. Personally I'm loving my Vertex 4, but even the regular Marvell controllers look to be a lot more flexible. One SandForce drive is just like another, because SandForce writes the firmware. Probably because the firmware is where all the dedup and compression code is. And yet people spew venom not at SandForce but at OCZ, who admittedly made a mistake by joining the "me too" SandForce brigade, but come on no venom for other SandForce makers?
Completely agree. People direct their hate toward the wrong company. That being said, the problems seem to be fixed. Sucks to be an early adopter but sometimes that's the price you're going to pay when you're at the bleeding edge.
Interestingly enough I bought a Vertex 3 the first week it was out and had no problems. Still running on the original FW. Go figure.
OCZ needs to partner with HGST (formerly IBM's hard drive unit) and JMicron to create the worst SSD known to man. It will be known as "DEATHSTAR SS, STUTTER EDITION"
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
27 Comments
Back to Article
damianrobertjones - Friday, May 11, 2012 - link
...Money didn't stop the fast progression of tech."As expected, there is a decline in performance when moving from a sixteen package design to an eight package design."
Paul Tarnowski - Sunday, May 13, 2012 - link
Looking at recent prices I am amazed at the number of SSDs that hover around $100 CA for 128 GB after rebate. Mind you, the ones that I'm thinking about are from ADATA, who seem to have been making a big push recently, but other drives are starting to mimick their pricing.LB-ID - Monday, May 14, 2012 - link
OCZ still has a lot to prove to me before I'd ever touch one of their products again. Too long a history of buggy performance and crappy, blame-the-customer-first support.Bateluer - Monday, May 21, 2012 - link
90% of the time, the customer is likely the problem, sadly.Pessimism - Wednesday, May 16, 2012 - link
I specifically boycott products that are only competitively priced through use of a rebate. I understand the financial side of using a rebate vs a sale price from the vendor's perspective, but until they make great strides in making the process less hostile towards the customer (forms "lost", rejected over insignificant errors or imaginary reasons, 8 week processing times etc) they are an absolute deal-breaker for me.Give me an automated system that accepts an online form, optically recognizes a scanned image of the product UPC code, and accepts a scanned image of the receipt to be audited at random, which INSTANTLY transfers funds to you through paypal or interac email money transfer and we will talk.
<end rant>
zcat - Friday, May 11, 2012 - link
Looks like the SSD Bench section is missing quite a few other current drives besides just the Vertex 4's:- Samsung 830 (the "best" mlc)
- Crucial m4 128G (one of the most popular @ $125, though slower on writes)
- Intel 520
- Kingston HyperX
- Mushkin chronos
- Sandisk
Kristian Vättö - Friday, May 11, 2012 - link
Samsung 830 is in the bench, including all retail capacities (64GB, 128GB, 256GB and 512GB).128GB Crucial M4 is also in the bench: http://www.anandtech.com/bench/Product/425
60GB and 240GB Intel 520s are in the bench.
240GB Kingston HyperX and HyperX 3K are in the bench.
We don't have any SSDs from Mushkin yet, hence no data in the bench. However, due to the popularity of Mushkin, I'll try to get my hands on their Chronos series ASAP, but it will still take awhile.
We'll have a 480GB SanDisk drive in the house in a week or two. However, keep in mind that all SF-2281 based SSDs are pretty much the same, thus we don't test every SF-2281 based SSDs that's out there.
Also, keep in mind that some models may be missing some tests, i.e. we only ran the most essential tests and left some out to save time.
zcat - Friday, May 11, 2012 - link
Ah - yes - thanks for clearing up why I thought some drives were missing. The default 'Anandtech Storage Bench' just had the fewest drives, and I lazily ctrl-f searched the page ala cpu passmarks searchable list.zcat - Friday, May 11, 2012 - link
One more thought:Perhaps add a new (default) benchmark that is simply the average bench of the intersection of all benches & drives?
Kristian Vättö - Friday, May 11, 2012 - link
I agree. Right now it defaults to 2010 Light Workload - Average IOPS test, which is something we don't even use anymore. I sent Anand a message about this and will let you know what's his take on this.mfed3 - Friday, May 11, 2012 - link
i love how each revision of ssds is getting worse and worse to cut costs. this never happens in any other computer component. greedy ass companies.SamLJG - Friday, May 11, 2012 - link
What is getting worse exactly?Taft12 - Friday, May 11, 2012 - link
What? That happens for EVERY computer component! Have you seen how awful today's low-price laptops are? And rotational hard drives, printers, power supplies, ....It's not just computer components of course, this applies to everything sold at Walmart and most other stores for that matter. Crap quality for cheap price.
bupkus - Friday, May 11, 2012 - link
One can only prey that China doesn't take up the production of toilet paper and condoms.Pessimism - Wednesday, May 16, 2012 - link
http://www.intimategadgets.com/product/smooth-natu...Electronically Testeo.... for your safety.
Kristian Vättö - Friday, May 11, 2012 - link
I don't really get what you mean. If you are referring to the fact that the 64GB drive uses 8 NAND packages instead of 16, that's the case with all 64GB SSDs with 2Xnm NAND. 2Xnm MLC NAND die is 8GB, so that limits the maximum amount of NAND packages to 8 in a 64GB drive.KungFu_Toe - Friday, May 11, 2012 - link
OCZ has a release candidate for firmware revision 1.4. Apparently, it gives a serious boost to write speeds. I'd really like to see these drives retested and benchmarks for 64 GB version with this firmware.Kristian Vättö - Friday, May 11, 2012 - link
We'll most likely wait for the release version before publishing any data. I talked with Anand earlier today and he said he has done tests with the 1.4 RC but there are still a few things that should be fixed in the final version, if I interpreted his messages right.StormyParis - Friday, May 11, 2012 - link
A French site has been tracking faulty returns to a big French e-tailers for a few years. SSDs in general are not very reliable, and OCZ SSDs in particuliar are crap:http://www.hardware.fr/articles/862-7/ssd.html
Return rates:
- Crucial 0,82% (last year: 0,8%)
- Intel 1,73% ( 0,1%)
- Corsair 2,93% ( 2,9%)
- OCZ 7,03% ( 4,2%)
I woldn't buy that brand.
Taft12 - Friday, May 11, 2012 - link
The fact that OCZ is that much worse than Corsair which is selling the IDENTICAL PRODUCT (Sandforce controller plus some NAND on a PCB) is really damning. Just awful. Yuck!Kristian Vättö - Friday, May 11, 2012 - link
While I'm not saying that the data is totally irrelevant, I would like to point out a couple of points.1. OCZ Vertex 2 is experiencing the highest return rate but it's no longer a high performing SSD in today's standards. The return rate of Vertex 3 is noticeably lower, although still fairly high.
2. OCZ was the first one to use SandForce's second generation (SF-2000 series) controllers and they had exclusivity for months. Some of the returns are due to early firmware flaws and bugs, which were fixed before others started selling SF-2000 SSDs.
3. There is no breakdown of Corsair's faulty SSDs. One should note that Corsair sells Marvell based SSDs as well, so their SandForce based SSDs could have a high failure rate but their Marvell SSDs would balance the return rate.
4. There is no info on return rate of Indilinx Everest based SSDs, which is what Vertex 4 is. Now OCZ has more freedom when it comes to firmware and they are not at the mercy of SandForce. OCZ has also said that they are spending more time and money on validation now.
5. It includes only four brands so we don't really get a good idea if OCZ's return rate is due to SandForce or is the problem at OCZ's end.
Again, just pointing these things out. However, I don't think that data is completely valid when talking about Vertex 4.
npaladin2000 - Friday, May 11, 2012 - link
How dare you use logic and reason against someone's righteous anti-OCZ crusade??? :)Seriously, I would tend to agree. I remember when the original Vertex and Agility drives came out, people loved them. They went downhill when they started selling SandForce drives, but Vertex4, Petrol, and Octane are non-Sandforce. Personally I'm loving my Vertex 4, but even the regular Marvell controllers look to be a lot more flexible. One SandForce drive is just like another, because SandForce writes the firmware. Probably because the firmware is where all the dedup and compression code is. And yet people spew venom not at SandForce but at OCZ, who admittedly made a mistake by joining the "me too" SandForce brigade, but come on no venom for other SandForce makers?
josephjpeters - Sunday, May 13, 2012 - link
Completely agree. People direct their hate toward the wrong company. That being said, the problems seem to be fixed. Sucks to be an early adopter but sometimes that's the price you're going to pay when you're at the bleeding edge.Interestingly enough I bought a Vertex 3 the first week it was out and had no problems. Still running on the original FW. Go figure.
Pessimism - Wednesday, May 16, 2012 - link
OCZ needs to partner with HGST (formerly IBM's hard drive unit) and JMicron to create the worst SSD known to man. It will be known as"DEATHSTAR SS, STUTTER EDITION"
Klinky1984 - Saturday, May 12, 2012 - link
Hard drives really are not much more reliable:http://storagemojo.com/2007/02/19/googles-disk-fai...
Look through any newegg hard drive review section and you'll find people putting reviews up like "bought 4, one was dead" or "died after 3 months".
SSDs I think will eventually exceed hard drives as far as integrity goes, some are already doing so.
iwod - Friday, May 11, 2012 - link
Am i correct in thinking the spec table isn't updated with the latest performance improvement from v1.4 ?And if the 64GB version would get similar performance benefits as well or is that set in stone already?
andrebrait - Friday, May 11, 2012 - link
Didn't OCZ release a firmware update which nearly doubled the write performance of these SSDs?