The OCZ Vertex Series SSD 预览，性能超猛

yym

The OCZ Vertex Series SSD

There isn't really a whole lot you can do to "pretty up" a solid state drive...but it's what is on the inside the counts.

For our testing we received a 120GB model of the Vertex series though there will also be 250GB and 60GB models available when the finally hit the street.  You can read on the features list that the Vertex includes a 64MB on-drive cache, just like the OCZ Summit series we previewed earlier.  

The Vertex uses the now-familiar 2.5-in form factor used in notebook drives.  I know many of you have been looking, and you should know that OCZ also has some of the 2.5-in to 3.5-in adapters for standard chassis.  

The Vertex also uses the standard SATA data and power connections so you should be able to use this drive anywhere a 2.5-in drive exists.  

The inside of the drive is at least more interesting than the exterior!  Here you can see eight Samsung flash memories - these are only single stacked since we are testing a 120GB model rather than 250GB.

Flipping the PCB over we see another set of eight Samsung chips as well as two more interesting offerings: the Indilinx Barefoot controller chip and a 64MB chip for cache from Elpida.  

We also think we see something else important on these drives: a flush-mount capacitor that in theory should hold enough power to allow the controller to flush the 64MB of cache into non-volatile flash memory in the event of an unexpected system shutdown.  This is important in order to prevent any data loss while data is in the interim step between cache and flash.

Taking a closer look at the Indilinx IDX110 chip clearly reveals it to be an ARM processor design with special logic as we described on the previous page.

Now let's get to the real meat of the matter - how does this new player in the SSD field perform in our tests?

yym

Testing Methodology and System SetupTesting Methods

Our storage testing hardware remains unchanged and is more than capable of putting the stress on these standard and solid state drives necessary for proper benchmarking.  We are using a Core 2 Duo E6400 processor on a Gigabyte P35 motherboard that uses the ICH9 south bridge.  The OS of choice is Windows Vista x64 SP1.  The rest of the components are listed in the table below.

Our tests are a good mix of synthetic and real-world benchmarks.  The two PCMark suites offer a good combination of both worlds while the IOMeter, HDTach, HDTune, Yapt and our custom File Copy test round out the selection to cover just about all bases.  If you have any questions about our tests just drop into the Storage Forum and we'll help you out!

Test System Setup

We are certainly getting a nice assortment of solid state drives for performance comparisons - the new Vertex SSD we are testing today will go against quite a few of OCZ's other options including the original SATA 3.0 Gb/s 64GB drive, the Summit 250GB and the Apex 250GB SSD.  We of course will continue to pit these drives against the Intel X25-M 80GB SSD and the fastest standard spindle-based hard drive, the Western Digital VelociRaptor 300GB.

Hard Drive Test System Setup
CPU	Intel Core 2 Duo E6400
Motherboard	Gigabyte P35-DS3R
Memory	OCZ Technology 2 x 2GB DDR2-800
Hard Drive	OCZ Vertex 120GB SSD OCZ Summit 250GB SSD OCZ Apex 250GB SSD Intel X25-M 80GB SSD OCZ 64GB SSD SATA-II Western Digital VelociRaptor 300GB WD3000GLFS Western Digital Raptor 150GB
Sound Card	none
Video Card	ATI Radeon X800 XL
Video Drivers	Catalyst 8.3
Power Supply	Antec TruePower 480 watt
DirectX Version	DX10
Operating System	Windows Vista x64 SP1

• PCMark05
• PCMark Vantage
• Yapt
• ATTO Disk Benchmark
• IOMeter
• HDTach
• HDTune
• PCPer File Copy Test

PCPer File Copy Test

PC Perspective File-Copy Test v0.3

Our custom PCPer-FC test does some fairly simple file creation and copy routines in order to test the storage system for speed.  The script creates a set of files of varying sizes, times the creation process, then copies the same files to another partition on the same hard drive and times the copy process as well.  There are four file patterns that we used to try and find any strong or weak points in the hardware: 10 files @ 1000 MB each, 100 files @ 100 MB each, 500 files @ 10 MB each and 1000 files at 1 MB each.  

Hmmm....these are very interesting initial results; the OCZ Vertex 120GB drive is the first drive to consistently beat out the Intel X25-M 80GB and also bested the OCZ Summit drive.

yym

HDTach 3.0.4.0

HDTach v3.0.4.0  

HD Tach will test the sequential read, random access and interface burst speeds of your attached storage device (hard drive, flash drive, removable drive, etc). All drive technologies such as SCSI, IDE/ATA, 1394, USB, SATA and RAID are supported. Test results from HD Tach can be used to confirm manufacturer specs, analyze your system for proper performance, and compare your performance with others. HD Tach is very easy to use, quick, and presents data in easy to read graphs, including the ability to compare two storage devices on screen at the same time for easy analysis.

The burst rate seen by our HDTach benchmark is somewhat confusing as it is considerably lower than the other scores from SSDs we have tested.  Even more interesting: the burst speed is exceed by both the sustained read and write speeds.  The issue is likely in how the cache on the drive is utilized; either way it's the 223 MB/s read and 199 MB/s write speeds that really impress us.  
HDTune tests a similar level of features that HDTach does, but with a slightly different access pattern and thus can provide us with an additional set of benchmark numbers to compare between storage configurations.  Here we can get the minimum, maximum and average transfer rates as well as the burst rates, access times and CPU utilizations.

Once again the performance in our HDTune 3.50 benchmark impresses with the OCZ Vertex 120GB SSD and is basically on par with the Intel X25-M.

yym

ATTO Bench

ATTO Disk Benchmark 2.34  

Though we have been using ATTO for a lot of our recent SSD testing, we didn't have results from earlier drives but I felt the results seen here were worth discussing.  The Intel X25-M is still the top drive for read performance - it not only peaks higher but it ramps up faster as our transfer size increases.  For writes though, the addition of the 64MB of cache with the Indilinx Barefoot controller clearly gives the OCX Vertex drive the best overall performance.  

PCMark05

PCMark05 v1.1.0  

For these tests, we use RankDisk, an application developed and copyrighted by Intel. In our testing, we found RankDisk to be suitable for a neutral benchmark. RankDisk is used to record a trace of disk activity during usage of typical applications. These traces can then be replayed to measure the performance of disk operations for that usage.
RankDisk records disk access events using the device drivers and bypasses the file system and the operating system's cache. This makes the measurement independent of the file system overhead or the current state of the operating system. In replaying traces, RankDisk always creates and operates on a new dummy file. This file is created in the same (or closest possible) physical location of the target hard disk. This allows the replaying of traces to be safe (does not destroy any existing files) and comparable across different systems. Due to the natural fragmentation of hard disks over time, they should be defragmented before running these tests.
The traces used for each test were created from real usage. The traces contain different amount of writing and reading on the disk; total ratio in the HDD test suite disk operations is 53% reads and 47% of writes.
The following input traces are used:
Windows XP Startup: This is the Windows XP start trace, which contains disk activities occurring at operating system start-up. The test is 90% reading and 10% writes. This trace contains no user activity.
Application Loading: This is a trace containing disk activities from loading various applications. It includes opening and closing of the following applications:
Microsoft® Word
Adobe® Acrobat® Reader 5
Windows® Media Player
3DMark®2001SE
Leadtek® Winfast® DVD
Mozilla Internet Browser
The application loading trace is 83% reads and 17% writes.
General Hard Disk Drive Usage: This trace contains disk activities from using several common applications.
These are:
Opening a Microsoft® Word document, performing grammar check, saving and closing
Compression and decompression using Winzip
Encrypting and decrypting a file using PowerCrypt
Scanning files for viruses using F-Secure® Antivirus.
Playing an MP3 file with Winamp
Playing a WAV file with Winamp
Playing a DivX video using DivX codec and Windows® Media Player
Playing a WMV video file using Windows® Media Player
Viewing pictures using Windows® Picture Viewer
Browsing the internet using Microsoft® Internet Explorer
Loading, playing and exiting a game using Ubisoft Tom Clancy's Ghost Recon
The General Usage trace is 60% reads and 40% writes.
Virus Scanning: Virus scanning is a critical task in today's PC usage. As the major bottleneck of scanning viruses is in hard disk activity, it is reasonable to include virus scanning as a HDD test. The test consists of HDD activity of scanning 600MB of files for viruses. The Virus Scanning test is mostly disk reading (99.5%).
File Write: This trace contains disk activities from writing 680MB files on the hard disk and no read operations are involved in this test.
Disk idle times have been compressed to 50 milliseconds to speed up the playback time. Our studies showed that 50 milliseconds was the smallest idle time interval that didn't affect the test results. The results of the HDD tests are reported in Megabytes processed per second.

Our PCMark05 results were bit of a mix - we actually have gotten in a pair of these Vertex drives, each with different firmware.  In our early testing we have seen the two different drives react quite differently with PCMark05 (and PCMark Vantage) - while one firmware would do better in our OS tests, the other would accell at the application tests.  What you are seeing here in our results is a composite of those scores; its because of that we are calling this article a "preview" as we expect to see one more firmware release before the retail release.  

yym

In a similar fashion to our PCMark05 results, the Vantage test saw some fluctuation between different firmware revisions.  Obviously there are some questions then in regards to these benchmarks that we will be addressing in the full, retail release review.  

IOMeter v2006.07.27 - IOps

IOMeter v2006.07.27  

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.

Somewhat disappointingly, our IOMeter results with the OCZ Vertex 120GB drive were much lower than we expected.  They were, however, consistent across the two firmware revisions we had in the lab.  In no case is the Vertex really a competitor with the Intel X25-M and even the OCZ Summit series of drives outperforms the Vertex in the one test where results looked "typical".

yym

IOMeter - Average Transaction Time

Our results look pretty much the same with the the average transaction time as they did when viewed as IOs / second - the OCZ Vertex drive really only in competition with the Summit or X25-M in our web server results.  

Yapt v0.3

Yapt v0.3  

Yapt (yet another performance test) is a benchmark recommended to me by a pair of drive manufacturers and was incredibly difficult to locate as it hasn't been updated or used in quite some time.  That doesn't make it irrelevant by any means though, as the benchmark is quite useful.  It creates a test file of about 100 MB in size and runs both random and sequential read and write tests with it while changing the data I/O size in the process.  The results are a good look at overall drive performance.

For both random and sequential reads, the OCZ Vertex drive is one of the top performers in our selections of competitors here - it is able to compete with the likes of the OCZ Summit and Intel X25-M drives.  Write speeds are much less impressive though not completely off the charts in terms of standard drive performance.

yym

Performance over time

As with our OCZ Summit preview article, we are going to attempt to look at long term SSD performance with the Vertex in a somewhat compressed manner.  

Our first HDTach run - 223 MB/s read and 199 MB/s write

Here you can see the results reported in our preview - you can clearly see that already there is some "jumping" associated with SSDs occurring half-way through the test.  

After our round of tests - 222 MB/s and 136 MB/s

After our entire onslaught of tests this is what the HDTach results look like; you can see that there has been a drop in write performance even though read performance is still very good.

A full pass with HDTach results in a stable drive of 223 MB/s read and 218 MB/s write

Our "fix" for this type of performance drop is to put the drive through a "full" benchmark with HDTach that writes an entire pass across the drive.  The result is a "fresh" drive with performance hitting near its theoretical maximum.  

A "fresh" run of ATTO

Though HDTach did see dramatic write performance decreases with a "used" drive, is the same true for another test like ATTO?  Here is our initial ATTO result screenshot...

ATTO on a "beat up drive" - very similar performance, not a lot of degradation

Follwed by a run just after a completely random IOMeter pass for about an hour on the drive.  While performance did fall just a bit (look at the early results for each run) the results are much more manageable.

All of this means that our testing with long term performance on SSDs is more valid and important as we move forward.  Expect some more detail and results on this phenomena in our full review.  

ConclusionsPerformance

The new OCZ Vertex solid state drive is one of the most impressive SSDs we have tested to date.  I know that there is much excitement in the hardware community about this drive and for good reason: it uses a brand new controller from a brand new company and as such performance was an unknown.  Many in the community have been billing it as the first true competitor to Intel's X25-M line of mainstream MLC drives and it indeed lives up to that status.

The Indilinx Barefoot controller, based on a custom logic ARM processor, is doing nothing dramatically different than JMicron or Samsung or even Intel controllers; but it has proved itself to be a strong competitor for a first offering.  In my initial preview testing the OCZ Vertex drive offers read speeds just slightly below that of the Intel X25-M 80GB SSD but in most cases has a much faster write speed than the Intel drive thanks in large part to the addition of 64MB of on-drive cache.

Interestingly, the OCZ Summit, based on a new Samsung controller chip, also uses 64MB of cache and is also a strong competitor to the Intel X25-M and OCZ Vertex SSDs.  There are still a few question marks including IOMeter performance and fluctuations in the PCMark05/Vantage arena that we hope to work out by the time the drives hit the streets this month.

Firmware Enhancements

As I mentioned on the PCMark05 and Vantage pages, there were some differences between the benchmark results of our two different Vertex drives with different firmware revisions.  Just to illustrate that difference, look at the screen shots below:

PCMark05 - first Vertex sample


PCMark05 - second Vertex sample

You can see there are some pretty big differences here: pay attention to the first three benchmark results as they see a dramatic difference going from the first drive we received to the second.  

Needless to say, OCZ is working on performance improvements with the Indilinx engineers and they will likely have a new firmware installed on the Vertex series before they are sent to the retail market.  That does mean they will be released even later than expected; but I think we would all rather have the drives running at top speed.  

Capacity and Pricing

While the OCZ Summit series won't be available until the end of March, the Vertex drive should have actually been for sale already.  The firmware updates are causing a bit of delay, but I would expect to see them for sale online as early as the end of February.  

The Vertex drives will come in 60GB, 120GB and 250GB variates with prices of $199, $399 and $829, respectively.  That actually puts the price per GB at an even $3.33; compared to the $/GB of the Intel X25-M (currently at $4.73) that is actually quite a good deal if performance on the Vertex lives up to what we have seen here.

Final Thoughts

The OCZ Vertex might be the most hyped up solid state drive yet to be released, thanks in large part to the new controller offering from Indilinx and the competitive price per GB.  As it turns out, the performance lives up to the excitement and expectations for the most part and we are hopeful that with one final firmware update OCZ will have a winner on its hands with the OCZ Vertex series.

yym

http://www.pcper.com/article.php?aid=670   转贴



性能值得 期待 ， 很是奇怪 ，这个文章 已经出来2天了， 国内 怎么还没 见到 转贴  ？


里面 英语 相对比较简单 ，有配图 ，大家将就看看吧。

superray

已经有人转了，
LZ不是早就预定了么，怎么还么发货？

yym

还没上市啊 ，据说月底  ，在最后 修正。我这2个月 订了 几次 都被 取消了，也等麻木了。

Duron750

等降价等降价

gxggxk

文化低  看不懂英文的说

breaf

{wink:]HOHO！

D65

下了班，还是看中文比较休闲...

thinkpig

yy版定的是什么价格？
打算几个怎么用？

kobe606

看上去还可以~30G,500考虑下

cha20000

不错 等价格合适

bluepig

固态硬盘要好卖 一个是价格 一个是卡

cdpfjp

晕，虽然不算文盲，过了个6级，看着也是天书