Anvil's Storage Utilities

[2f_tfe]

some charts of "Threaded I/O" benchmarks (made with RC1),
t would be wonderful if ASU could do it automatically, something such as
a batch procedure with variable "queue depth" (1,2,4 .. 128) ,
i think it will save a lot of time to reviewers and enthusiast users


Crucial m4 64GB (2x RAID0 - Intel Z68) Threaded Read I/O


Crucial m4 64GB (2x RAID0 - Intel Z68) Threaded Write I/O

(write back cache ON/OFF)


i noticed that with a single drive there's no improvement after queue depth of 32
(virg SSD and HDD) , but with RAID configurations there is a little improvement,
that's why i asked you about the decision to test only at mid-low queue depth
(for the primary benchmark)


just two things about the function of "copy to clipboard", i think it should do
a full report of results and not a brief summary, such as AS-SSD and CDM ,
and a new label (button/icon) could be useful , between menu "Screenshot" and "Help"


bye
luke

[Kain665]

Hey Anvil,

Any chance you can increase QDs to 256 on the "threaded" tests? It would also be nice if users could specify the number of threads used to drive the load and how long the test runs for. Would you consider this for a future update?

[Kain665]

Also, what is the functionality of the greyed out IOmeter tab?

[Anvil]

There will be a lot more in the full release

The IOmeter tab is a module for reading/analyzing IOmeter result files.

[Kain665]

Great!

When do you think the full release is coming?

[Anvil]

RC3 is now available.

Download link

Fixes

- Endurance testing, MD5 testfile can no longer be cached.

[Cooperdale]

This is by far the best disk testing utility out there. I hadn't used it since beta 9, and with the latest rc final score has decreased of about 10%. I was wondering if this can be attributed to different testing (or result-weighing) methods in new builds, or I might actually have a performance degradation to investigate.

[Ourasi]

Same ICH/PCH drivers?

Many of the latest months of Intel betas will drop scores in some configurations..

[Cooperdale]

Same ICH/PCH drivers?

Many of the latest months of Intel betas will drop scores in some configurations..

I have to double check, but I don't think I've updated any Intel drivers this year. Anyway the last test was taken about 2 months ago, I positively didn't change anything driver-wise since.

One thing I did was update Diskkeeper to the latest version.

[Ourasi]

One thing I did was update Diskkeeper to the latest version.

I hope you do not let Diskkeeper get anywhere near your SSD's.. If your sig is for real regarding the SSD model in your box, I'd expect some degradation.

[Cooperdale]

I hope you do not let Diskkeeper get anywhere near your SSD's.. If your sig is for real regarding the SSD model in your box, I'd expect some degradation.

Actually I updated my sig, that was a very old rig, now I have two Corsair Force 3 in Raid0. Anyway, I've been using Diskeeper for at least 3 years with ssd drives and the performance never degraded, at least with Ocz drives. Was perfect for 6 months with the old version and the new drives. But you never know...

[-X-hellfire]

This is weird.... I thought last night about the documentary, that was on the telly 1 - 2 years ago, about a metal band called Anvil.

http://en.wikipedia.org/wiki/Anvil_%28band%29
http://www.imdb.com/title/tt1157605/


Today, after my computers raid crashed, I saw this thread about Anvil, and thought it was a thread about the metal band Anvil, but no it was the TS username.

I looked through the the thread and it looks like a very good thread that I will bookmark and remember later to test out the latest beta.


Sorry for the OT but how strange that I thought about Anvil last night

[B.A.T]

Looks like more and more review sites are using ASU

Tom's Hardware

Hardwarecanucks

Pureoverclock

Tweaktown

Thessdreview

[AMD_Freak]

Looks like more and more review sites are using ASU

Tom's Hardware

Hardwarecanucks

Pureoverclock

Tweaktown

Thessdreview

Ive been using ASU since early Betas for Reviews on PureOC can hardly wait for the Final

[Zaxx]

...can hardly wait for the Final

+1...but I'm soo glad he's taking the time to do a top notch job.

[Highendtoys]

The only thing I would like to see different is a 4K QD32 test in the standard set. It's common to have companies quote IOPS numbers from QD32 and it would make reviewing easier....and actually give me something to talk about with this test other than how much a SF based drive slows down with incompressible data.


Bill Gates is using it too at RWLabs.com as well.

[B Gates]

Of course I use it at RWL its great

[alan1476]

Of course I use it at RWL its great

And a smart man you are.

[dragmore]

Featurerequest. Add 64KB IO-size as part of the benchmark! (since thats rather standard of most semi new databases afaik)

br TE

[B Gates]

here's a new score to beat. just ran it.

[Anvil]

So, that's what can be expected from the GS

(great score even though it's 0-Fill)

@dragmore
...there will be more IO sizes.

[Zaxx]

Did they stick with the 2282 with all 32 nand chips as with the Force 3 240GB? I'm guessing probably not.

[B Gates]

Did they stick with the 2282 with all 32 nand chips as with the Force 3 240GB? I'm guessing probably not.

no its a 3/4 board with a 2281 and 16 chips

[One_Hertz]

Anvil - any luck changing the read tests to test actual data that is on the SSD instead of the test file?

[B Gates]

So, that's what can be expected from the GS

(great score even though it's 0-Fill)

@dragmore
...there will be more IO sizes.

Okay, How about this: