Anand Vertex 3 Review

Ok this graph really grinds my gears:

Image removed because Anand fixed the issue

Looking at all the other Vertex3 reviews, the ~190MB/s values for 4KB reads are actually the result from using QD32 NOT QD3 like Anand is claiming. Extrapolating from all the other reviews, the real QD3 numbers should be around half of the result he claims, if not less. By definition, the numbers from QD1 can not increase more than 3 times for QD3. Since QD1 is 3XMB/s or less, QD3 can not possibly be 190MB/s.

Does anyone here talk to the guy? Could you drop him an email? He is going to keep reusing this most likely false info (like he did in the Intel 510 review).

Most likely a typo in the header of his chart is all. He just left off the 2.

Nobody tests with QD=3 that I know of.

You can email him yourself. Click on his name in the article for a mailto link. :)
And yeah, I think Anand's results are highly suspicious. And it's not just V3...

@ RyderOCZ: Anand does. He reviewed Intel 510 and tested it @ QD=3 as well.

Oh really.. interesting.

One-Hertz, I will test if for you right now if you like.

Only SATAIII board I have up and running is an MSI 890FXA-GD70 though.

A quick iometer run would be awesome :up:

It looks suspicious.

Here's from my previous testing of some of those SSDs.
Based on the C300 the readings look like they are measured at QD3.
The C300 at QD32 ~ 220MB/s

http://www.xtremesystems.org/forums/...9&d=1270601186

Ok, I am apparently rusty with Iometer... where do you set the QD? {embarassed}!!

http://www.xtremesystems.org/forums/...8&d=1294496006

Outstanding I/Os is what you are looking for.

edit:

Link to thread with iometer profiles

That MIGHT also be compressible data?

Quote:

---

Originally Posted by alfaunits

That MIGHT also be compressible data?

---

This should not matter much for low QD 4kb random tests. The bottleneck in this case is the time the SSD takes to locate each piece of data which is requested (i.e. access time). Compression can not speed up this kind of latency. Compression can speed up how quickly the data is read once it is located.

This should show my config for the test and the results.

My system and drivers:

http://www.ocztechnologyforum.com/st...ex3/system.jpg

Iometer config:

http://www.ocztechnologyforum.com/st...3/iometer1.jpg
http://www.ocztechnologyforum.com/st...3/iometer2.jpg
http://www.ocztechnologyforum.com/st...3/iometer3.jpg

Result:

http://www.ocztechnologyforum.com/st...ter_result.jpg

Thanks Ryder!

This proves Anand made a mistake in his testing...

Or he just put the QD32 results with the QD3 header :)

He did use P67 and I used SB850, but I can't see it making 100+ MB/s difference.

I just repeated the test with QD32 and got ~160MB/s average.

Note that my tests are not with a fresh SE'd drive, I have Windows on it (for another machine) so all the blocks are not fresh and clean.

Quote:

---

Originally Posted by RyderOCZ

Or he just put the QD32 results with the QD3 header :)

He did use P67 and I used SB850, but I can't see it making 100+ MB/s difference.

I just repeated the test with QD32 and got ~160MB/s average.

Note that my tests are not with a fresh SE'd drive, I have Windows on it (for another machine) so all the blocks are not fresh and clean.

---

Putting QD32 results against other drives which were ran at QD3 is a mistake... Hopefully he fixes it.

Yes, Vertex3 can definitely do more on a better controller, fresh drive, etc. Around 70Mb/s probably.

Someone call the Shrimp, to make sure.
Frankly, I believe it is a case of a good controller (ICH) and a fresh drive with compressible data. If the data consists of typical IOMeter data (a repeat of 0123456789 over and over), the controller could very easily push that in a very small amount of space and need not read the actual cell content but minor auxiliary data that says the LBA is a copy of another LBA which was already read (hence in cache). 128MB cache or even 64MB would be enough to hold an 8GB IOMeter file in compressed form.
Either way the results would be pointless, but the test might be true.

That mistake does make a huge difference to that chart.

The SB850 is actually quite alright at low QDs, not so sure it will make that much of a difference.

The MyCE review puts the V3 at ~95MB/s at QD4 and pcper shows that it really starts to pick up speed at QD > 4

Wow - this changes the game a bit. This drive is not so head n shoulders over the 510 as we may have thought?

OK, I hope he reads his e-mails. These 4K numbers in Vertex 3/Pro and Intel 510 reviews are nuts...

well i read vertex 3 reviews last night until my eyes bled. i mean i read like 9 of them front to back. i too noticed some inconsistencies in testing, and some sites just obviously do not know how to test.
the anand results did look skewed and that is a helluva call there OneHertz, nice catch.:up::up:


the Anand result and the results of some other sites had me scratching my head.:confused:

but lets keep things in perspective here: from all of this reading that i did i came to one conclusion

this is one helluva drive. i think right now it is the clear leader (duh) and unless intel really tweaks its next series that is coming out, this here Vertex 3 might be the leader of the class for a good good while.

i wasnt reading reviews over and over without the intentions of picking up a few (or 8) of them very soon.

you might think that would saturate most raid cards, but ohhhhh...what i know....:p:

now all we need is a lower capacity V3 to save me at least a little bit of money :yepp:

@anvil: hey there is a 2008 verison of IOMETER, but i cant seem to find it anywhere. is this the same as the beta that you posted in the other thread?
ryder: can we please get a QD1 test?

Quote:

---

Originally Posted by Computurd

ryder: can we please get a QD1 test?

---

By praz:

http://www.xtremesystems.org/forums/...1&d=1299070784

For some reason, the 4KB random read QD=1 numbers are hugely different between AS-SSD and and CDM on Vertex 3. I asked about that in the AS-SSD thread on OCZ forums, hoping the AS-SSD author might have some ideas, but no replies yet.

Here is an example of the discrepancy:

http://thessdreview.com/our-reviews/...-and-win7-wei/

The QD=1 results I have seen from IOMeter are closer to AS-SSD than to CDM. I think CDM's number may be misleading.

yeah i would like to see iometer QD1.

btw F@32 is that jim carrey?

Quote:

---

Originally Posted by Computurd

btw F@32 is that jim carrey?

---

Lol yeah, he was epic in the SNL. Especially Black Swan skit: http://www.hulu.com/watch/205721/sat...ive-black-swan

[Anandtech] Correction: OCZ Vertex 3 Random Read Performance Data

Anand read your email :)

Quote:

---

A huge thanks goes out to AnandTech reader Andrei and some very attentive Xtreme Systems forum members. I just got an email pointing me to this thread where one particular number from our OCZ Vertex 3 and Intel SSD 510 articles was called into question. The problem? The 4KB random read numbers for the Vertex 3 were supiciously high. The reality? They were incorrect.

I was just alerted to the error and quickly powered up the SSD testbed to recreate the test. It looks like the original numbers were either run at a queue depth of 32 or accidentally copied from one of the runs of 4KB random write tests. Either way the number was incorrect and has been fixed in all affected articles....

---

http://images.anandtech.com/graphs/graph4202/35666.png
http://www.anandtech.com/show/4204/c...rformance-data

Good to hear. Who actually ended up emailing Anand? I didn't...

And my 70mb/s prediction was almost spot on (:

you got some recognition, and you deserve it :) that was very confusing for many im sure, made it seem like so much conflicting info out there. man think of how many thousands of people have read that....

Awesome stuff guys - great catch. Please pardon my ignorance, but where are 4K QD-3 random reads typically in action on a normal user's system? Is it the boot sequence, application loads, anything and everything? As opposed to QD32 or QD1... Or rather, how important is this metric to begin with?

Looks like the V3 is a bit lackluster in this particular area. I guess I'm looking for as much justification for purchasing two of these things as I can find. :)

There we go. Looks like C300 is still ahead. :p:

Quote:

---

Originally Posted by zalbard

There we go. Looks like C300 is still ahead. :p:

---

What's up with that anyways? I keep reading that the 510 controller is a "newer" C300 controller. Why the huge performance difference?

Quote:

---

Originally Posted by One_Hertz

Good to hear. Who actually ended up emailing Anand? I didn't...

And my 70mb/s prediction was almost spot on (:

---

Wow, he was fast. E-mailed him last night.

Quote:

---

Originally Posted by chuchnit

What's up with that anyways? I keep reading that the 510 controller is a "newer" C300 controller. Why the huge performance difference?

---

People are saying it's firmware. I personally think there is also some hardware difference, either a somewhat different chip, or something else... :shrug:

Quote:

---

Originally Posted by Brahmzy

Please pardon my ignorance, but where are 4K QD-3 random reads typically in action on a normal user's system? Is it the boot sequence, application loads, anything and everything? As opposed to QD32 or QD1... Or rather, how important is this metric to begin with?

---

Unfortunately almost no tasks use QD>1 on a desktop. Ao1 and others did awesome job in this thread analyzing it: http://www.xtremesystems.org/forums/...d.php?t=260956

It is sickening that new gen SSD serving popular crowd a coolaid of increased sequential performance...

I'd wait gor G3 25nm, aka 52x series from Intel before jumping the gun on 510 or V3.

Quote:

---

Originally Posted by alfaunits

Someone call the Shrimp,

---

you called? :D
:rofl:

Holy crap! Zalbard got banned?

Quote:

---

Originally Posted by Computurd

Holy crap! Zalbard got banned?

---

Sam_Oslo too :confused::confused::confused::confused::confused:

wow. i heard that zalbard is a casualty of the news section. gotta be careful over there...but the rules do keep the site from devolving into a bunch of flaming fanboys i guess. not to say that is what zalbard or sam_oslo were doing but i do see the need for the rules.


sorry for the off topic :)

Quote:

---

Originally Posted by prava

Sam_Oslo too :confused::confused::confused::confused::confused:

---

That is not a suprise. He got banned here in Norway too @ diskusjon.no forums :yepp:

wow. BanHammers all over the planet!

"It takes two to tango" as they say.

(The ban Nizzen refers to was years ago, jfyi)

there were like 4-6 people banned in a single thread

back on topic, the QD=3 shows the vertex 3 behind the c300, does this mean the old stuff is still just as incredible?

and why does no one review with real tests anymore. i know anand is trying to make their own benchmark suite, but id like to see realistic game/app loads and bootup times.
also i think someone should throw it into a PS3 and see the differences there too, lol. i know im going to be putting my classic Agility in there when i upgrade my desktop.

It does look like the V3 is a bit behind at low QD.

As you correctly are referring to, it's getting a bit boring with all this QD1 talk, of course it is important but it's like buying a car with the gear-stick locked in 1st gear.

Real-life benchmarks are hard to find, PCMV is OK but it has it's flaws.

edit:
The old stuff is still great, I do believe that some of the speed gains we are seing is directly related to the incredible sequential speeds we are seing.

by old stuff i mean c300, not as old as my horribly out dated Agility

too many tests just show off the max sequential speed.
if you really are doing video work, a large traditional raid will offer the same speed, for the same price, and give you 1000x more space. so im not a real fan of PCMV or other tests like that.

i wonder if what we need first is an extensive survey of what needs people filled when buying an SSD, so reviewers can get a hint.

i did see one review with an SSD thrown into a console, (just so happen to be an Agility, and it just so happen to be a game i play, GT5), and load times for maps were dropped nearly in half across the board, and load times for menus were all dropped aswell. people who play that game will know how much it sucks to move around and how long the wait can be (half cause of the PSN, and half cause of the disk)

Quote:

---

The old stuff is still great, I do believe that some of the speed gains we are seing is directly related to the incredible sequential speeds we are seing

---

excellent point. even i get blinded by the 4k qd 1 jazz...and i shouldnt the sequential is very very important as well.

70 percent in gaming usage.

^^
I was referring to C300s, Vertex 2s, G2,... :)

Finding a common type of benchmark is almost impossible, PCMV is a bit of everything but it's getting dated.

Creating such a benchmark on a forum like this?, I don't know, benchmarking is actually a lot of work.
A survey would be OK, it would at least show some sort of demand for benchmarks outside the usual ones.

@CT
We need both random and sequentials, your new testrig should make it possible to solve some of these questions :)

Quote:

---

Originally Posted by Computurd

excellent point. even i get blinded by the 4k qd 1 jazz...and i shouldnt the sequential is very very important as well.

70 percent in gaming usage.

---

Sequential is there, but looking at http://www.xtremesystems.org/forums/...&postcount=159 it never goes above 100MB/s :confused:

I'd say that it depends on the computer as such and the level of OC, a number of factors.

4K QD1 is key for most tasks, but I think there is a ceiling at which this performance metric can be utilised.

I can’t really explain this well, or even be sure it is correct, but it seems that priorities are set within the OS I/O stack to suit the poor performance of HDD.
As an example when I boot I have to manually interact with a program as it loads. Only after I have done this will the next app load. It the OS I/O stack overloaded a HDD with multiple concurrent tasks the HDD would seize up. It doesn't seem to be SSD aware.

The other issue I have not really looked at yet is the impact of sequential I/O’s that are interspersed with small random reads. This is what appears to happen with game playing.

I guess also that there are plenty of other factors like the CPU & GPU will also have a big impact.

I don’t really know......for discussion really.

Anyone else having issues with this iometer Fileserver pattern performed by TomsHardware Link

Vertex 3 is great but having 55.000-60.000 iops at QD1, i think not, unless there are a lot of workers, which in turn means that QD needs to be multiplied by # of workers.

I haven't found their definition of the fileserver pattern but originally (as defined by Intel) it is as follows

Attachment 112645

Attachment 112646

edit:
pcper performed the fileserver benchmark and it looks correct. Link to pcper
Attachment 112647

Argh. March 21st cannot come quick enough!

Quote:

---

Originally Posted by Anvil

Anyone else having issues with this iometer Fileserver pattern performed by TomsHardware Link

---

60K IOPS at QD=1 is hard to believe. But it seems there is something different with the Vertex 3 at QD=1. Different from other SSDs. I'm not sure if it will lead to high performance in real usage, but look at CDM 4KB read, 35-37MB/s for Vertex 3. That is higher than I have seen for any single SSD. Of course, AS-SSD only reads about 20MB/s for Vertex 3.

So there is something about the Vertex 3 that can result in some benchmarks measuring very high QD=1 values in some situations. I have no guesses yet what that may be.

Quote:

---

Originally Posted by johnw

60K IOPS at QD=1 is hard to believe. But it seems there is something different with the Vertex 3 at QD=1. Different from other SSDs. I'm not sure if it will lead to high performance in real usage, but look at CDM 4KB read, 35-37MB/s for Vertex 3. That is higher than I have seen for any single SSD. Of course, AS-SSD only reads about 20MB/s for Vertex 3.

So there is something about the Vertex 3 that can result in some benchmarks measuring very high QD=1 values in some situations. I have no guesses yet what that may be.

---

What where the test sizes on CDM? Where they low 50MB-100MB ones? And how much onboard cache does the Vertex 3 have?

The test sizes are whatever people used. Just look at the Vertex 3 reviews on the web. Most of them appear to have used the default 1000MB setting for the latest CDM version.

I agree Anvil. one thing that is becoming very very clear from reading the scores of Vertex 3 reviews is that most of these reviewers really dont know wth they are doing. not all, but some. some are very good though, and i do like seeing their different approaches, but some of the testing borders on retardation LOL
its cool to keep it simple for people, and sometimes they can be too technical. that is fine. but being just plain wrong or not knowing how to test is another thing.

thats why i had to read so many though. ive seen so many wild inconsistencies over the course of like 9 reviews that i had to kind of piece them together to get the 'real' picture.

Quote:

---

Originally Posted by Computurd

I agree Anvil. one thing that is becoming very very clear from reading the scores of Vertex 3 reviews is that most of these reviewers really dont know wth they are doing. not all, but some. some are very good though, and i do like seeing their different approaches, but some of the testing borders on retardation LOL
its cool to keep it simple for people, and sometimes they can be too technical. that is fine. but being just plain wrong or not knowing how to test is another thing.

thats why i had to read so many though. ive seen so many wild inconsistencies over the course of like 9 reviews that i had to kind of piece them together to get the 'real' picture.

---

I haven't bothered reading any of the reviews. These quick reviews tell you very little and it seems that direction is given on how to test to show the drives in the best light. Give it a few months and the picture will become clear from end users on this forum.

Hey guys been missing for a week and wont be on much as im studying programming as my main unit which is hard, Anyhow is there a g3 review out as of yet ? Also i've been reading the c300 is still top dog for random 4k speeds, whats the go with that ?