Areca 1880

[PnoT]

WOW, Nice results CT!

Btw, which drives are you using in that array?

[Computurd]

eight gen 1 vertex R0
they seem to be limping along nicely...

[lowfat]

108)also pcie packets (set at 1048 will go to 4096 though)

What do you mean by this? I've never heard of such a thing.

[Computurd]

it is the size of the pci-e packets that each device sends across the pci-e bus. when the devices on the bus communicate, they do it by sending "packets". It is much like traffic...so when they are sharing the bus each device sends its packet, then the next, so on, so on. rapid fire. but each packet is a certain size. now you increase the size thus increasing the amount of info sent across on each of that particular devices "turn". in some motherboards you can set the size of the packets sent, from 128 (which is default on most boards) then 512, 1094, 2048, 4096.
only thing is if you set the packet size too high you will introduce stuttering into your video, if it is also coming across pci-e express ports. the videos packets will not be being recieved fast enough for good play, etc. Uber tweaking at its very finest, finding the good settings yield great results. shhh! dont tell anyone the secret i still havent done it yet on this run btw, havent had time yet for those last few percentage tweaks, jumping hurdles right now

Jumped up 5k points on PCMV only at 4.3 on my 920, i am up to 28k\
i need a 980x
http://service.futuremark.com/compare?pcmv=358183

[Tiltevros]

nice results paul. on monday i will try to play with PCM i want my name on that list . maybe wont be the top cause of the MP but im hopping to be on top 20 somewhere :S

[Computurd]

ty tilt, there is more to be given, having issues with either bad fanouts (gotta spare though) or a bad drive(spare for that too). will have to do more testing in the AM to isolate the problem. slowing me down big time. You should receive great results with those dual xeons, and the benefit of the dual IOH!
need SB these guys cpu scores are tremendous! cant wait to see what nizzen/anvil can do with their gear, some awesome scores inbound i am sure!

[Tiltevros]

i want to try them on 5520 chipset to see how ARC works there and what will be the benefits of stock settings no Overclock for me

[Ao1]

it is the size of the pci-e packets that each device sends across the pci-e bus. when the devices on the bus communicate, they do it by sending "packets". It is much like traffic...so when they are sharing the bus each device sends its packet, then the next, so on, so on. rapid fire......

Hey Comp, great results. Looks like the Areca 1880 is shaping up very nicely. Not wishing to derail the thread but the issue you raise on shared bandwidth on the pcie interface was one of the reasons I was excited about the Viking nand solution for a “ram” based SSD module. By all accounts it can send data via the memory socket and that gives a direct interface to the cpu. A memory socket seems to be the fastest interface on a mobo as far as I can see with nanosecond latency and transfer speeds up to 17,066 MB/s, but maybe I am missing something as it got slated in the news thread when it was announced.

[Computurd]

yes i couldnt agree more, there are some issues with sharing of the bus, and the thing about the packet optimizations, is that if you go too far you will end up adding latency, so there is a definite inherent weakness of the interface. not neccesarily the lanes themselves, but that they terminate at the IOH, which in turn handles traffic. some of the serious advances with sandy bridge should be absolutely phenomenal, such as the inclusion of the pic-e controller onto the chip, should handle the throughput much faster, not to mention pci-e 3.0. the pci-e 3.0 controllres are where my next goal lies. 600,000 iops and a dual 1.4 ghz ROC on the next gen of the LSI. hopefully areca is going to implement this same controller into their line of products. i think now that they are over the whole marvell fiasco you should see much faster r&d in the future, which is an impression i have gotten from speaking with them.
the nand module you are speaking of is an excellent concept, and i am sure that one day we will see some implementation of that, as a matter of fact things are headed that way with memresistor (am i saying that right?) which could double as storage and ram at the same time, exciting stuff! i could be wrong, but i believe that the implementation you are speaking of is actually only using the power form the memory socket to power the device, yet it has a sata port on the side, and that is why they arent too excited about it, it still connects via the sata port i think. future versions will go over the memory lanes themselves. they do seem like very good ideas though. its crazy that a year ago everyone was thinking that getting the nand on the pcie was the way to go...but things are advancing so fast, now they are shifting to memory/ram implementations over the ram slots!

my testing continues today, only have made about three solid runs, having issues with my system...

[Spoiler]

....not neccesarily the lanes themselves, but that they terminate at the IOH, which in turn handles traffic. some of the serious advances with sandy bridge should be absolutely phenomenal, such as the inclusion of the pic-e controller onto the chip, should handle the throughput much faster, not to mention pci-e 3.0.....

You can try your theory out on a P55 motherboard. The P55 has the PCI-e controller on the cpu. I'd love to see the differences in latency from a P55 and X58 using a high end raid controller or even better an IOxtreme card. I mentioned this a while ago, but I haven't seen anyone draw any comparisons yet.

[Computurd]

that is an excellent observation. i will not be testing the p55 though, unfortunately. or maybe i will...they are pretty cheap. An IOextreme isnt going to really test the throughput as well as the high end raid cards out right now. it simply does not have the sheer IOPS.

[Computurd]

actually, steve-o arent you beating on a P55 right now? are there any differences with your 1231 that you could relay? or is it about the same?

[SteveRo]

Haven't even tried to see if the board would boot with a 1231 -it is a very little board - only one pcie slot and no onboard raid.
I will post a pic.

[Computurd]

ty sir...

also on some testing of a different variety, after reading this form gfody at storage review.com:

In my testing with Areca the storport driver doesn't scale past 32 where the scsiport scales up to 256. The storport tests showed significantly more iops than the scsiport at the same queue depth though. Sorry I don't have the data handy.

i ran some more tests, and he is correct the 4k scales much higher, up to 482 mb/s with QD of 258. going to make up another graph to compare, but that will be in the future..still doing some pcmv action.

[Computurd]

wow the 1231 prices just went down..

[SteveRo]

Ok so way off topic but CT asked.
Gigabyte H55N-USB3 mini ITX under cascade.
Almost as good an overclocker as the x58 UD7.
I did need to do a silver pen mod to the 655K for CBB but other than that the mobo/proc is stock.

The H55N-USB3 looks to be a very little board expecially whne setting on the big cascade.

[Computurd]

jesus lord that is beautiful MAN it is small...your evap dwarfs it...poor little board...
it would be very hard to get that card on there...

[SteveRo]

Thanks CT!
How's the pcmv progress coming - you know to use x64 of the app right?

[Computurd]

The PCMV work continues, but i fear my ol 920 is just at the limits my man...here is a view of my cramped work area...need a bigger room but wife wont have it!
OF course some new firmware for the 9260 released today while im busy with other things....really need to get that on to try, with LSI new firmwares are UBER....

[SteveRo]

Good morning CT,

Looks like you have all kinds of room!
Yes 980 really helps pcmv suit scores no question.
You do have a couple excellent storage options between the 1880-4GB and the 9260, keep pressing and good luck!

[Computurd]

yea, believe me i wont stop till i see smoke!

[SteveRo]

All right Mr CT - you are making progress!
Looks like you just posted the #9 pcmv score - 28124 - very nice.
You have the highest quad cpu pcmv score on the ORB - looking good!
The 8 in front of you are all 980's - time for a new proc soon?
Very nice work and much congrats!

[SteveRo]

@ Mr Peanutskeeter - nice score 24769 with 980 at 4763?
#12 on the Orb!
What are you using for storage?

[Computurd]

just hit 29437 steve....still working my friend...finally put all ten vertex on the thing so here we go........dialing up the clocks a bit too..
EDIT:29796

[SteveRo]

All right! - #7 and only ~1k back from Mr Nizzen!
and with a 920 - well done!
If you went to a six core you would add 2+K to that score.

edit - also - very nice AS SSD score too!