ATTO Disk Benchmark is perhaps one of the oldest benchmarks going and is definitely the main staple for manufacturer performance specifications. ATTO uses RAW or compressible data and, for our benchmarks, we use a set length of 256mb and test both the read and write performance of various transfer sizes ranging from 0.5 to 8192kb. Manufacturers prefer this method of testing as it deals with raw (compressible) data rather than random (includes incompressible data) which, although more realistic, results in lower performance results. The 512GB result is on the left with 256GB on the right:
Specifications list 535MB/s read and 475MB/s write for the 512GB version and 535MB/s read and 380MB/s write for the 256GB version which is in line with what we are seeing here. This was a bit unexpected simply because ATTO favors the use of highly compressible data in its testing and we expected a performance drop. Adjusting the Queue Depth in ATTO testing shows no change in performance.
CRYSTAL DISK BENCHMARK VER. 3.0 X64
Crystal Disk Benchmark is used to measure read and write performance through sampling of raw (0/1 Fill/compressible) or random data which is, for the most part, incompressible.For our testing today, we will concentrate on testing with incompressible data samples and the 512GB will be shown on the left with 256GB on the right.
Our Crystal Diskmark results typically fall below that of ATTO which is normal but we thought the high sequential and 512k read results of the 512GB drive were a bit low. What was really nice to see was a low 4k score above 100MB/s on the 512GB Vertex 4.
Up until recently, AS SSD was the only benchmark created specifically for SSD testing and it uses incompressible data. AS SSD, for the most part, gives us the worst case scenario in SSD transfer speeds while using SandForce Driven SSDs as they use compression in storage as discussed earlier. Many enthusiasts like to benchmark with AS SSD for their needs.
V4 512GB SSD
This is perhaps the best result that I have seen in AS SSD and definitely shows the strengths of the Vertex 4 512GB SSD. Every aspect that a reviewer looks for is taken care of here; great high sequentials, dynamite low 4k speeds, equal and high IOPS and great disk access time. This result is something that enthusiasts have been in search for since SSDs joined the consumer storage scene.
Nice numbers but I’ll be waiting until after the first firmware rev comes out before diving in. The 512Gb pricing is very attractive and makes the V3 480 price look pretty ridiculous. I’ll be curious how these bad boys handle RAID 0.
You’ll find loads of benchies here…;)
https://www.ocztechnologyforum.com/forum/showthread.php?100402-Vertex-4-performance-and-analysis
It’s a barn burner for sure…
Nice benchmarks… bit of tweaking on those benches as well. Did I see a 150+ 4k random write?
Great numbers, i think imma pick one up
https://goo.gl/G2vxa
“OCZ Technology is standing by the Vertex 4 with a five year warranty”
“…the fact that the Vertex 4 has an industry leading four year warranty.
So is it four or five?
It’s five, but either way 5 isn’t industry leading with Intel,Plextor, and OWC beating them to the 5 year warranty by a long time. This was surely a reaction to the upmarket brands offering two more years of coverage than everyone else.
So it’s good, but it’s not industry-leading… that would be 5 years and 1 day of coverage ;).
Dooooh. My bad and fixed! It was a late night.
Give the fw a few months and throw in some 24nm toggle and watch me pounce. 4 sho
I like the specs on the 512GB version very much. If we can see above 550MB/s read and write on future releases then I will be truly impressed.
Les,
Have you heard of anyone working on a benchmark or test suite for SSDs in virtualized environments? Maybe VMware Workstation for the enthusiasts up to the vSphere 5 for the heavy-hitters. The increased write performance in the V4s should make these very attractive in such environments.
Not yet although I can say we are always working with companies behind the scenes to improve our test environment!
I consider low performance on compressible data a bug and I hope OCZ will fix it with upcoming firmwares. If it’s not, this SSD isn’t so good. I don’t understand how it can perform so bad in PCMark. Compressible and incompressible data are both important for user.
Hey Les, LOVE the site. I’m an amateur PC enthusiast (26 years old) and recently landed a killer job and I’m looking to blow my first few paychecks on an ultimate system. What type of SSD would you recommend? I was looking at the 256GB Revodrive 3 X2 (RVD3MIX2-FHPX4-240G, listed as 230,000 IOPS on newegg)… but would a couple of these bad boys raided together be better? I’m planning on building the system in May to coincide with the release of Diablo 3, the GTX 690, and possibly an Ivy bridge CPU.. (However, I like the X79 platform so I might stick with a 3930 for now…)
Any recommendations? I’d love to hear your thoughts on 2x SSDs in Raid 0 vs a killer PCIE SSD.
The 3×2 definitely gives you the power you are looking for except it adds on about 10-15 seconds at initial boot. Myself, if I were going to RAID two SSDs right now, I might also consider the new Vertex 4, however, many would shy away simply because it is the new kid on the block.
Gimme a price point that you want to stay within for this part of the build.
The 3×2 will definitely giev you the power you are looking for but it also adds a bit of time at boot for its bios. If I were to look at two ssds in a RAID config, I would seriously look at the new Vertex 4 but many might shy away because it is the new kid on the block…
Tell you what…what is the abs max you want to spend on this part of your system?
Oh and thank you for the compliment! Nice to hear I am not the only reading my words eheh.
I’d like to keep the total system below $3.5k… The 3×2 on newegg.ca was $829 and I guess I’d be okay with that if the performance advantage was justified. I’m only going to go with 1x GTX 690, and a 3930 instead of a 3960, so my budget isn’t exactly limitless, but I’m trying to gain enough knowledge in the meantime so that come May (when I buy/build my PC) I’ll get pretty much the best of everything without spending much more than $800 on a SDD.
It seems like any way you slice it, a PCIE SSD (at least in the $800 range) is going to be faster/better and also cheaper than 2x raid SSD. I’ve heard about the boot time, I’m not really concerned with that. I’m also hoping that the super talents will be awesome and under $1k as well.
I know for my needs, I’m just mostly going to be gaming/playing Diablo 3, I won’t need anywhere near a GTX690 or a 3×2 SDD lol 😛 But its still awesome, and I’m kindof leaning towards a pci-e just because of the awesome/cool factor. Sortof the same reason why I’m leaning towards the X79 as opposed to the new Ivy Bridge… quad channel memory + pcie 3.0 vs a “tick” in intel’s cycle… just feels like they’re going backwards. But I’ll make that decision in May….
E-mail sent.
Are these good enough for SMB/SMEs, not just home enthusiasts? There is a mantra that anyone using non-Enterprise class SSDs in a 24/7 environment should be shot. But now I’m not so sure! I’m seriously considering populating a QNAP TS-879U-RP iSCSI SAN with 8 x Vertex 4s (6 in RAID10 and 2 as hot spares). This will be for a vSphere 5 VDI environment – very intense random workloads and roughly 60/40 reads to writes. I expect these drives to easily last 3-5 years, so even if I void the warranty by using them in an enterprise scenario, who cares? At this prices I can buy 3 for the price of a decent “enterprise-class” SSD. I really hope someone does some enterprise-class benchmarking (24/7 over maybe a week) soon.
Usually for deploying consumer SSDs into these environments adopters wait a period of time to ensure that there is a high level of reliability. The performance isn’t as important as the reliability. Since this SSD uses a proven Marvell controller, it should be fine, but should is a big word. Personally I would wait until the next firmware comes out to smooth out a few release errata before making a large investment. There will be the need for Overprovisioning to help keep performance steady. With a 20% OP these drives will probably remain very solid for a good period of time.
Utilizing RAID 10 you are wisely mitigating any risk, so I feel that you would be taking a safely calculated risk.
Hey guys. Check out my OCZ Vertex 4 speed test on YouTube. It shows my home PC’s boot up time, then I load a VDI environment running on Citrix XenApp and VMware Workstation.
https://youtu.be/YrnIcudM7zo
I’d welcome any comments, feedback or questions.