New LSI 9200 series controllers: 6Gb/s, 2.88 GB/s seq. reads, 1.87 GB/s seq. writes

[Computurd]

LOL yes it is not challenged with these drives!

[alfaunits]

lol
just how fast is your "futuristic" lsi @ real apps ?
http://www.youtube.com/watch?v=MvBKB2NndbI

WOW! STARTING 53 APPS AT ONCE! WHAT REAL WORLD ... Oh wait..
BS.

[Biker]

I'm with the apples to oranges argument.

The 1231ML does not compare to the 9260 -8i because one is SATA only and the other is SAS.

The areca card excels at low latency (4k is faster) and the LSI provides high sequential bandwidth (more than 4 drives = unrestricted throughput).

I only got the LSI card because my current cards would not work with my X58 board.

If there was a SATA only PCI-e 2.0 spec card on the market at the time I would have grabbed it in a heartbeat

[Computurd]

The card cant 'excel' in one small category (4k)and by and large be better, not possible. especially considering the real world usage. The bootup i/o meter file is about the closest thing to real world OS use, and i have yet to see one beat the LSI....also you must take into consideration that, according to my testing small file speeds went up 20 percent with the last firmware release, this card is new new new...so much room to improve expand and grow, those other cards are done. What about when trim is implemented? that is going to change the landscape entirely. It adds more instructions, will old gen be able to keep up with the new gen? Not hardly!

[lowfat]

it is from the OCZ site, here is the link to the thread itself.. where you will find that pattern near the bottom of the first post
http://www.ocztechnologyforum.com/fo...ad.php?t=54752
I would love to see your results! I can guarantee that those acards will rape the VERTEX themselves...maybe...but maybe not the controller! What kind of controller do you use?

fixed. btw you image tagged the url

[Serra]

lol

just how fast is your "futuristic" lsi @ real apps ?

http://www.youtube.com/watch?v=MvBKB2NndbI



you guys still cant comprehend specs/benches/real apps

Napalm...

Real-world benches are cool and all, but you really can't post a video like that and say "OMG MY LAST-GEN CARD >>>> NEXT GEN CARDS BECAUSE MY CUSTOM APPLICATION THAT IS HEAVILY CPU AND SSD DEPENDANT DID GUD!!1".

Yeah, you went fast. Zoomjmygodwafflehouse! But how the hell do we test against that? To properly compare cards we would have to find the same quantity of the same non-cheap SSD's and run the exact same test on the exact same hardware, because otherwise we're comparing apples to oranges. Clearly, that's not going to happen.

This is just like when you used to post some (I want to say video encoding) times and said it proved how great your array was compared to anyone elses. You straight-up challenged me on it in one thread, despite the fact that your processor was around 2x as fast as mine and the test was heavily CPU dependant.

Real-world perforamance is great. Don't get me wrong about that. But FFS either make a repeatable test that is provably not CPU constrained and release it for general use or STFU. Is it really so hard for you to believe that new hardware released after lots of time spent on R&D by a company much larger than Areca could outperform your magic hardware?

I'm not really convinced Areca will release any new stuff anytime soon. I have no hard data to back that up as it's just a feeling, but something inside me tells me it may be a couple of years really...

Areca is not the holdup on new cards - it's Intel. Areca does not make it's own RAID chips, they use Intel's. And if you think Intel is developing SATA 3 SSD's and not working on a next-gen RAID chip you're bonkers.

[Computurd]

OH thx LOL

[Computurd]

and i would also like to point out that by no means is his hardware performing so much better than anyone else's because of some sort of skill or mastery of the art of RAID>>....anyone could buy the same stuff and do the same thing~!
and STILL get beat by a 6 GB/s card!

[NapalmV5]

lol as soon as i started posting again here.. so much hating..

so far none of you posted anything real apps of your future controller.. lol you guys crack me up

you guys got problem with a 106 app loadup ?? it just goes to show what the whole system can do @ real apps.. cpu/ram/os/controller/ssd all contribute

your future controller should blow away my setup easily (benches).. yet when it comes to real apps none of you dare post such



looks like i should take an even longer break from this section.. but ill be back when i get the napalm wrath/decimation done

your future controllers will beg for mercy!!

[Computurd]

Ah no one is really hating here, we are just asking for you to post something up that has real value, something that will show us what your STORAGE system/controller can really do! You are the one who makes the challenges, but then refuses to step forward! Have a nice day

[Computurd]

Well videos are nice and all mr. napalm, but the only way for you to conclusively compare your RAID system to ours is not via videos or anything else, but by simple small file transfers, we are not racing CPU's here...you say your raid card is faster. Prove it. And not with a video, but widely accepted benchmarking to give us a feel of the speed of your system. You claim that your storage system will beat anything out there, and you scoff at this next gen controller. However the top three raid card manufacturers are all building their next raid cards around THIS VERY CHIP that is on this controller. Now, apparently you know more than all of these people combined, because yours is so much better? LSI, INTEL, and ARECA themselves are all wasting time making controllers that are inferior to yours???/?LOL
why would areca even bother going next gen then?
Dude the arc-1680ix blows that arc-1231 outta the water, yet i have one setting in a box simply because it cannot even begin to compete with this LSI card here.
Run a bootup i/o meter or something significant.

[XS Janus]

Arguing about which benchmark holds any real value for real life performance difference is moot.

I like what Napalm showed us so far in his vids and in his reasoning. Real life apps can only benefit so much from higher sequential rights. And Napalm has a good point when he states that nothing has been shown that proved the new LSi card are better in that regard than his 4xSSD and Ar1231 setup.
The way I see it, the only way anybody can compare 2 cards realistically in real world situations is to have all the card you wish to compare and run them with same SSDs and on the same machine.

So, since Napalm has gone through all the trouble to try and show the performance of his setup, the best way to end this would be for him to try the same tests on an LSi card.
As a personal note I have a feeling he might be right. The difference in access time due to the lack of SAS overhead could very well be the key here. Sure he won't get any top scores on sustained side, but hey he's only got 4 SSDs as it is and if he thought adding more SSDs to his array would help in real apps, he would have already bought them! Just my opinion!

[Serra]

lol as soon as i started posting again here.. so much hating..

so far none of you posted anything real apps of your future controller.. lol you guys crack me up

you guys got problem with a 106 app loadup ?? it just goes to show what the whole system can do @ real apps.. cpu/ram/os/controller/ssd all contribute

your future controller should blow away my setup easily (benches).. yet when it comes to real apps none of you dare post such



looks like i should take an even longer break from this section.. but ill be back when i get the napalm wrath/decimation done

your future controllers will beg for mercy!!

You want to compare them? Ok. Downclock your CPU to stock so that others can potentially match it (you're at what, a 4.xGHz OC on a C2Q, right?), then make a full list of all programs that your batch file launches as well as their configurations. Next make sure that we have the ability to install them for free, because if they all cost money this is a ridiculous argument. Then, and only then, can we actually try to "race" you.

Hell, I'm on a laptop but I could make a batch file to open 1000 programs in 18 seconds, Napalm. # of programs/opening per second is an asinine and ridiculous test, because it obviously depends on WHAT those applications are and, yes, whether any CPU utilization is required to open them. Opening app's is not simply a matter of storage speed.

[Computurd]

Yes that is the issue that is being taken with this....There is a reason there that he is not showing us his real stuff. He needs to come forth with REPRODUCABLE results. Something tangible that can be proven. Right now we have a video that he uses to claim that he is the 'king' of the storage mountain. We have shown him what we have, but he will not show us, other than the vid of course. There are reasons that the people who build these things test them with programs that MEASURE things.
It is not about who is fastest to me, but obviously it is for him.
so he should prove it. He is the one who keeps making these claims. He cant and thats why he WONT.
HE has done nothing to prove to ME that his is faster than this card.
And really how do you think that sustained throughput has nothing to do with it when you are loading say 50 apps in seconds??? what do you think all of that is random read? The sustained read has a tremendous amount to do with it!!!
maybe if he wouldnt come in nutswinging with his vid then people wouldnt call him on it. so to napalm i say... or get off the pot! Run some i/o meters and post them up!

[Speederlander]

lol as soon as i started posting again here.. so much hating..

so far none of you posted anything real apps of your future controller.. lol you guys crack me up

you guys got problem with a 106 app loadup ?? it just goes to show what the whole system can do @ real apps.. cpu/ram/os/controller/ssd all contribute

your future controller should blow away my setup easily (benches).. yet when it comes to real apps none of you dare post such

looks like i should take an even longer break from this section.. but ill be back when i get the napalm wrath/decimation done

your future controllers will beg for mercy!!

Ok, I'm confused. Didn't you not that long ago post up the typical sequential benches (HDTach/tune) like they were the last word, refusing other benches like iometer, but now you won't accept benches of any sort as indicative of performance?

I tend to agree with Serra that you need to downclock to stock CPU/ram speeds. Otherwise why not just post heavily CPU dependent bench numbers for your "storage" while the system is coked up on LN2?

Either way, a good range of synthetic benchmarks can give a good feel for system capability. Post up tach, tune, iometer, crystal, etc.

You're talking big, but you need to meet people halfway and also post the standard benchmarks. Video alone doesn't cut it.

[Computurd]

Nice exit happy you found the door!

[Nizzen]

What can LSI 9200 series do to Pcmark Vantage?

Can it beat my 24/7 computer? I hope so

http://www.futuremark.com/community/...pcmarkvantage/

Areca 1680ix - 2gb + 5x Vertex

[Computurd]

i will do a run nizzen and we will see what we can get, now that is something that we can measure! will be fun to see how it turns out! thanks for the chance to give us something to compare it with!

[lowfat]

What can LSI 9200 series do to Pcmark Vantage?

Can it beat my 24/7 computer? I hope so

http://www.futuremark.com/community/...pcmarkvantage/

Areca 1680ix - 2gb + 5x Vertex

I'm number 17 and I didn't even use 4 drives for that run

[Computurd]

Nice lowfat! acards i take it?

[lowfat]

Nice lowfat! acards i take it?

No. Just two X25-Ms

[Ashraf]

NapalmV5, take a chill pill. No one is hating.

[Computurd]

ok so i just ran the HDD test suite on the vantage, but i ran that one only, it gave me a score of 35134 how does that stack up?

[NeedMoMegaHurtZ]

ok so i just ran the HDD test suite on the vantage, but i ran that one only, it gave me a score of 35134 how does that stack up?

I think this is old result, would have to ask Nizzen
http://www.xtremesystems.org/forums/...80#post3768780

[Nizzen]

This result is old yes.

You have to come over 100k points in hdd suite to be "cool"

Any results for lsi soon?

Edit: in pcmark vantage