Review Index:
Feedback

OCZ Vertex 450 256GB SSD Full Review - Indilinx drives 20nm flash

Subject: Storage
Manufacturer: OCZ Technology
Tagged: ocz, vertex 450, ssd, 20nm, sata

IOMeter v2006.07.27 - IOps

Iometer is an I/O subsystem measurement and characterization tool for single and clustered systems. It was originally developed by the Intel Corporation and announced at the Intel Developers Forum (IDF) on February 17, 1998 - since then it got wide spread within the industry.

Meanwhile Intel has discontinued to work on Iometer and it was given to the Open Source Development Lab (OSDL). In November 2001, a project was registered at SourceForge.net and an initial drop was provided. Since the relaunch in February 2003, the project is driven by an international group of individuals who are continuesly improving, porting and extend the product.

View Full Size

View Full Size

View Full Size

View Full Size

Light desktop usage sees QD figures between 1 and 4. Heavy / power user loads run at 8 and higher. Most SSD's are not capable of effectively handling anything higher than QD=32, which explains the plateaus.

Regarding why we use this test as opposed to single-tasker tests like 4KB random reads or 4KB random writes, well, computers are just not single taskers. Writes take place at the same time as reads. We call this mixed-mode testing, and while a given SSD comes with side-of-box specs that boast what it can do while being a uni-tasker, the tests above tend to paint a very different picture.

Ok, now for the meat and potatoes. The Vertex 450 handles these mixed workloads with no issue, turning in the same excellent performance seen by the Vector. This is goot do see, as the added latency of moving to 20nm flash typically slows an SSD down a little (as can be seen when comparing the Vertex 3 (25nm) to the Vertex 3.20 (20nm)).

May 23, 2013 | 02:14 PM - Posted by derz

Good to see that OCZ is staying competitive in both price and performance/reliability.

I wonder if SATA 4 is in the works. Its just the SATA 3 interface holding SSDs back and PCI-E is the enthusiasts interface, while Thunderbolt isn't enticing in terms of cost.

May 23, 2013 | 10:10 PM - Posted by east side (not verified)

Sorry, but with SSDs like the M500 on the market you are seriously saying this is a good SSD? wow. It is terrible, a huge step backwards.

June 19, 2013 | 01:43 PM - Posted by Allyn Malventano

...so should I just quote other reviewers that openly state the M500 is a "middle-of-the-pack" drive? The only thing it has on the competition is its availability in a high capacity (960GB), and at that price, supply is so constained that in nearly a month there are only 7 newegg reviews. Oh, and it's still out of stock. Bonus.

That quote is from the same review site you claim to be superior further down in the posts here. Perhaps you should actually read their reviews instead of randomly fluffing and bashing.

May 23, 2013 | 10:13 PM - Posted by east side (not verified)

LOL you are using PCMark 5? wow. and Iometer from 2007? Someone tell Marty to get back in the Delorean and come to the present time! Cmon Mcfly!

May 29, 2013 | 11:11 AM - Posted by Allyn Malventano

Iometer 2007 writes data that is less compressible compared to newer builds. We stuck with 2007 because newer versions were giving an excessively unfair advantage to SandForce controllers. The most recent 1.1.0 RC1 added a selection box for data compressibility, but the results have proven inconsistent in my testing and I believe it's not ready for production use (it's an RC after all). This is the result of weeks / months of my own research covering several platforms and dozens of different SSD models.

Newer versions of PCMark are running similar / slightly tweaked traces and add a bunch of extra results that tend to confuse people more than the necessary result of just figuring out which drives perform better than others. Alternatively, some sites will just publish the "storage suite" figure of the newer PCMark versions (a *single number*), to avoid such confusion. I feel that dumbs down the results too far to be meaningful. PCMark 05 gives us a few *good* results of sufficient granularity to evaluate drive performance. We use Vantage for SSD caching tests mainly because it hits the drive with IO's over a larger test file, which is more representative of OS / application launches which would need to be cached.

If it ain't broke, don't fix it. I do not subscribe to the megapixel race of benchmarks. Graphics benches must be updated to support newer technology. The only new tech beyond AHCI for SSDs is TRIM, which *no* benchmarks support as it occurs on the kernel level, meaning that all of those shiney new benchmarks you see out there do nothing more than use different IO patterns that are not necessarily better or more representative of real world usage.

...and for the record, Mcfly ran circles around Biff.

June 8, 2013 | 08:39 PM - Posted by east side (not verified)

The 2010 versions of IOmeter are perfectly fine and used by all manufacturers and validation houses for testing purposes. Let me help you a bit here since your month of discovery didn't figure it out...it is only the Psuedo random that provides inconsistent results, and for a purpose. It is meant to test varying compressibility workloads.
The incompressible and fully compressible settings are just fine. The 100% compressible data is the exact same as the version that you are using. 2007 writes the same 100% compressible data, so claiming that the other versions give unrealistic results is ludicrous, you are using the very version that leans most favorably to SF controllers. Iometer 2010 has other architectural changes that also allow it to interface with modern drivers and operating systems in a more efficient manner. To claim that there 'no new tech other than ACHI' only illustrates your lack of complete knowledge of this subject. Disappointing.
PCM05 is outdated, period. There is a reason that benchmarking technology advances and improves, and that is to be more representative of the 'new normal'. How much uncompressed high definition (1080P, 4K) do you think they were doing in 2005? Has anyone even heard of F-Secure, the program used for virus testing in that bench? Do you really feel that this is representative of today's enhanced, newer, more efficient virus programs? It is a defunct program.
It also uses traces from 3dmark 2001. what were they testing back then, DOOM? Go back and read the blurb above your own results, and then tell us that those tests really have anything to do with today's usage. You are almost ten years behind. This is why your test results have lost relevance, and also why these reviews arent held in as high esteem as they were even a few years ago.
Leadtek® Winfast® DVD??? are you kidding me? This isn't about racing to get new benchmarks, or trying to keep up with the Joneses. This is about presenting relevant data to your readers to further their purchasing decisions, and this isn't doing it. Not even close.
Lazyness is what it is.

June 8, 2013 | 08:43 PM - Posted by east side (not verified)

"Alternatively, some sites will just publish the "storage suite" figure of the newer PCMark versions (a *single number*), to avoid such confusion."

What a ridiculous statement. There are a total of 6 traces in the HDD test for PCMark Vantage and PCMark 7. It outputs results in all of those categories. Have you even bothered to run these programs?
You are testing fresh out of box results on unconditioned SSDs, it is the most unrealistic results that you can possibly provide. This is like some joke from thessdreview.
Here, let me give you an example of real testing...
http://www.hardocp.com/article/2013/06/05/sandisk_240_gb_extreme_ii_ssd_...

June 19, 2013 | 01:33 PM - Posted by Allyn Malventano

Ok, so you showed me a page that takes a similar round of results and just spreads them out over 6 charts instead of my spreading those same results over two charts. More != better.

I've run PCMark Vantage and PCMark 7 on all SSDs tested since those benches were released. Vantage is good for demonstrating the effects of caching under Intel RST and SSHD configs, but that difference comes from the way the different benchmark lays down the test file.

All samples are preconditioned prior to testing, and conditioned again after operations that might TRIM areas of the drive. This is common sense. I had the long term conditioning and evaluation of SSDs figured out a long time ago: http://www.pcper.com/reviews/Storage/Long-term-performance-analysis-Intel-Mainstream-SSDs

P.S. I know who you are. Go astroturf elsewhere.

May 24, 2013 | 05:43 PM - Posted by OctaveanActually (not verified)

I'd give the Crucial M500 the nod too,....

I'd like to buy a 960GB model (~$600) some time soon. Definitely by the holiday season unless something faster, bigger and cheaper comes out.

The OCZ Vertex 450 specs look food though,..

May 25, 2013 | 03:47 AM - Posted by Anonymous (not verified)

SSD'S in general......I have both types of drives. I do video editing. Not that big of a differance in performance by the way.. The biggest let down was when ( no lectures about backing up please) One drive failed. The very expensive data salvage co. said everything was lost. "If it was a HDD, we could have retrieved the data off the platters". Make sure you back up everything if you want to use one of these drives. These drives, at this point, are a disapointment. Maybe after a few years I will check back.

May 25, 2013 | 05:32 AM - Posted by rrr (not verified)

And if you used only HDDs, you wouldn't have bothered with any backups? You sir, are asking for trouble.

Post new comment

The content of this field is kept private and will not be shown publicly.
  • Lines and paragraphs break automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd> <blockquote><p><br>
  • Web page addresses and e-mail addresses turn into links automatically.

More information about formatting options

By submitting this form, you accept the Mollom privacy policy.