| OCZ Vertex 3 Max IOPS Edition SSD |  |
| Reviews - Featured Reviews: Storage | |
| Written by Olin Coles | |
| Tuesday, 24 May 2011 | |
|
Page 5 of 13
SSD Testing MethodologySolid State Drives have traveled a long winding course to finally get where they are today. Up to this point in technology, there have been several key differences separating Solid State Drives from magnetic rotational Hard Disk Drives. While the DRAM-based buffer size on desktop HDDs has recently reached 64 MB and is ever-increasing, there is still a hefty delay in the initial response time. This is one key area in which flash-based Solid State Drives continually dominates because they lack moving parts to "get up to speed". However the benefits inherent to SSDs have traditionally fallen off once the throughput begins, even though data reads or writes are executed at a high constant rate whereas the HDD tapers off in performance. This makes the average transaction speed of a SSD comparable to the data burst rate mentioned in HDD tests, albeit usually lower than the HDD's speed. Comparing a Solid State Disk to a standard Hard Disk Drives is always relative; even if you're comparing the fastest rotational spindle speeds. One is going to be many times faster in response (SSDs), while the other is usually going to have higher throughput bandwidth (HDDs). Additionally, there are certain factors which can affect the results of a test which we do our best to avoid. SSD Testing DisclaimerEarly on in our SSD coverage, Benchmark Reviews published an article which detailed Solid State Drive Benchmark Performance Testing. The research and discussion that went into producing that article changed the way we now test SSD products. Our previous perceptions of this technology were lost on one particular difference: the wear leveling algorithm that makes data a moving target. Without conclusive linear bandwidth testing or some other method of total-capacity testing, our previous performance results were rough estimates at best. Our test results were obtained after each SSD had been prepared using DISKPART or Sanitary Erase tools. As a word of caution, applications such as these offer immediate but temporary restoration of original 'pristine' performance levels. In our tests, we discovered that the maximum performance results (charted) would decay as subsequent tests were performed. SSDs attached to TRIM enabled Operating Systems will benefit from continuously refreshed performance, whereas older O/S's will require a garbage collection (GC) tool to avoid 'dirty NAND' performance degradation. It's critically important to understand that no software for the Microsoft Windows platform can accurately measure SSD performance in a comparable fashion. Synthetic benchmark tools such as HD Tach and PCMark are helpful indicators, but should not be considered the ultimate determining factor. That factor should be measured in actual user experience of real-world applications. Benchmark Reviews includes both bandwidth benchmarks and application speed tests to present a conclusive measurement of product performance. Test System
Storage Hardware TestedThe following storage hardware has been used in our benchmark performance testing, and may be included in portions of this article:
Test Tools
Test Results DisclaimerThis article utilizes benchmark software tools to produce operational IOPS performance and bandwidth speed results. Each test was conducted in a specific fashion, and repeated for all products. These test results are not comparable to any other benchmark application, neither on this website or another, regardless of similar IOPS or MB/s terminology in the scores. The test results in this project are only intended to be compared to the other test results conducted in identical fashion for this article.
|
|
Comments
The only "con" to your review, one of few I've ever had, is the slight hit you issued in your conclusion regarding "expensive enthusiast" product. OK, it is contrary to the masses and their hard earned dollars, no question. The 240 Gb version at $2.58/Gb is a goodly sum. But looking at it from a performance standpoint, and thinking of the terminology "enthusiast" makes me think of it not as a "con" but pretty mainstream, if not reasonable as far as bang-for-buck goes.
There's my two cents worth, raising the price to $2.60/Gb. But no one will pay, right?
80,103 IOPS Random Write (4k)
36,120 IOPS Random Read (4k)
556MB Sequential Write
515MB Sequential Read
Really solid build aswell, which perhaps isn't suprising since these guys apparently are better known in the military/rugged world. Shaping up to be one of best purchases I ever made!
The 240GB drive is much faster than the 120Gb drive. It gets even worse with 60/64 Gb(like vertex 2 series) drives which can be 70% slower than the 240Gb drive with incompressible write speed.
This review gives people the impression that a 120GB drive will perform like a 240GB drive but they do not.
I know its part of the marketing to send only the big fastest drives to reviewers but the performance difference between capacities should be more clearly explained
comment just sent - unclear as to why the above 2 fields have bounced back yet contents of message was sent?
Olin, my response was not directed at your valid comment.
I simply sent a lengthy comment to Benchmark Reviews about the relevance
of the performance of the slowest component in a system - the hard drive.
It was not published.
I run 2 older Sata2 SSD's (A-RAM 32Gb) in Raid 0 to kick start the system and run a few apps. When I was using a single SSD and Ubuntu 9.04 the read/write speed was 270/253 Mb. In terms of the 'IOPS' I had no idea and didn't appreciate the importance as an equal indicator of overall system performance.
All of my other mid range bits that make up the system are more than adequate for what I do....6 core AMD,a GTX460, a Raptor storage drive, a SATA3 & USB3 motherboard, blue-ray burner,2 LED 22" screens etc.
My point and I'm sure many of you have read that the money spent on the hard drive replacement 'SSD' should be the priority.
When I get the bucks OCZ will be my next upgrade.
Question : when in Raid 0, is the read/write & IOPS doubled also?
And, as Olin wisely adds, a LOT depends on the controller.
The closest we've come to near-linear scaling was the combination of a Highpoint RocketRAID 2720 and 4 x Hitachi 15,000 rpm 2.5" SAS HDDs each with 64MB of cache:
3 x HDDs in RAID 0 topped out at 441 MB/second
4 x HDDs in RAID 0 topped out at 618 MB/second
We would not have ever tried 3 x HDDs, except that one was
defective from the factory, so we configured a RAID 0 with 3
of these Hitachis, while we waited for the replacement to arrive.
But, this near linear scaling has been the exception, NOT the rule, in our experience with RAID 0 arrays (which we use often, primarily for speed).
Sincerely yours,
/s/ Paul A. Mitchell, Instructor,
Inventor and Systems Development Consultant
All Rights Reserved without Prejudice
Thanks again for another good review.
A question keeps coming up, as I read these various reviews:
if TRIM does not work with multiple SSDs configured in a RAID array,
do you happen to know if TRIM will work if each SSD is configured
in AHCI mode, and then joined to an OS "software RAID"?
Do you have enough instrumentation at your lab to answer this question?
I'm aware that an OS "software RAID" cannot host the OS,
but I have an XP software RAID running well with 2 x 6G WD HDDs,
hosting a dedicated data partition, and it appears to work just fine.
RSVP if/when you have the time.
Thanks again! Keep up the good work.
Sincerely yours,
/s/ Paul A. Mitchell, Instructor,
Inventor and Systems Development Consultant
All Rights Reserved without Prejudice
So long as the drives are in AHCI/IDE mode with an OS that supports TRIM (Microsoft Windows 7), then the commands will receive the SSD. Hardware RAID does not use AHCI/IDE mode, and will not pass along TRIM commands. So if you're using Windows 7 (not XP) and configure mirrored sets in the Disk Management console using AHCI/IDE SSDs, then it's not truly RAID and you'll have the benefit of TRIM support... with the penalty of a CPU/RAM hit for software mirroring.
Thanks again, Olin!
By "mirrored sets" do you mean RAID 1, for example, and not RAID 0?
Also, just to clarify, do you predict the same for RAID 0 / striped arrays using OS software RAID?
This could be the topic for another Review, perhaps.
p.s. With so many quad-core CPUs proliferating, now with hyper-threading too, the CPU-RAM hit does not seem to be a huge penalty to pay, particularly for desktop and workstation users.
Also, users who have shrunk their C: system partitions have probably also created dedicated data partitions e.g. D:+ Those dedicated data partitions would be good candidates for a software RAID, particularly if TRIM will work reliably, as described above.
Sincerely yours,
/s/ Paul A. Mitchell, Instructor,
Inventor and Systems Development Consultant
All Rights Reserved without Prejudice
XP/Pro supports software RAID 0 also:
we have it running on a P45 chipset
with 2 x 6G Western Digital 1TB HDDs.
I'm not currently a Windows 7 user,
but I would expect that the same is true
for Windows 7.
The partitions need to be changed to "Dynamic" first;
but after that, it's a piece o' cake to configure software RAID 0
as long as the drive letter is NOT C: .
One needs to use the OS to effect this configuration,
so structural changes to C: are not possible at this point.
Do you have some spare hardware laying around,
with which to try it? Maybe 2 x 6G SSDs (my preference)?
OCZ might be particularly interested in this experiment,
with your usual battery of benchmark tests.
You be the "Benchmark Man" -- to be sure!!
Sincerely yours,
/s/ Paul A. Mitchell, Instructor,
Inventor and Systems Development Consultant
All Rights Reserved without Prejudice
##softwaresecretweap ons.com/jspwiki/windows-7-software-raid-disk-partition-and-volume-context-menus
"New Striped Volume" is an option.
Sincerely yours,
/s/ Paul A. Mitchell, Instructor,
Inventor and Systems Development Consultant
All Rights Reserved without Prejudice
somehow "softwaresecretweapon s.com"
became "softwaresecretweap ons.com" above:
just eliminate that extra space after "weap"
Sincerely yours,
/s/ Paul A. Mitchell, Instructor,
Inventor and Systems Development Consultant
All Rights Reserved without Prejudice
The experiment I'm suggesting would need Windows 7
to configure a software RAID 0 with 2 x 6G SSDs
initialized in AHCI mode.
The question at hand: will Windows 7 pass TRIM commands
to 2 x SSDs configured in an OS software RAID 0?
Sincerely yours,
/s/ Paul A. Mitchell, Instructor,
Inventor and Systems Development Consultant
All Rights Reserved without Prejudice
[begin quote]
The Professional/Enterprise/Ultimate editions of Windows 7 officially support the following dynamic disk modes:
* Simple
* Spanned
* Striped (RAID-0)
* Mirrored (RAID-1)
These are the officially supported modes. It has been possible in previous Windows releases to enable unsupported modes through DLL modifications and/or registry changes, and this will probably be the case in Windows 7 as well. At your own risk, of course.
RAID-5 dynamic disks are only available in Windows Server editions.
Note that dynamic disks are not available in the Standard/Home editions of Windows 7.
[end quote]
Sincerely yours,
/s/ Paul A. Mitchell, Instructor,
Inventor and Systems Development Consultant
All Rights Reserved without Prejudice