nForce 790i SLI/Ultra chipset mobos - Reviews/OC/Guides

[noobzed]

my bandwith seem to be low, wonder why ?
corsair 1800c7
e8500 at 440 fsb and 1860 ddr3 7 7 7 20 1T rest on auto
p03 official bios

any idea ?

[dinos22]

my bandwith seem to be low, wonder why ?
corsair 1800c7
e8500 at 440 fsb and 1860 ddr3 7 7 7 20 1T rest on auto
p03 official bios

any idea ?

^^ that's why ^^

[noobzed]

okay
can someone pm me the un-offical P03R2 bios for my eVGA ?

edit: found this on evga website http://www.evga.com/forums/tm.asp?m=...1&key=&#310617
with this file: http://www.momoshare.com/file.php?fi...426289f325f84a
but it is a P01 with P1 and P2
and now bw is great, latency too:


edit: i am not sure it is a P01 cause it dated with 03/14 so...
but on boot screen, i can see P01 at the end of the bios name

[tekjunkie]

My problems with my board seems to be RAM related. I am getting memtest errors as I go over 360 FSB in sync, put in a RMA request @ newegg and am gonna return the RAM(lost 15% in restocking ). The 4 GB OCZ EB platinum that I ordered from mwave will get here tomorrow. Hope I will have better luck with this RAM than I had with my 2GB OCZ sticks. Thanks for all those who took time off to respond to my whining.

[tekjunkie]

I can't say this enough...
With 1T i run into a fsb wall @ 425mhz with my qx9650
With 2T i run into the wall @ 490mhz fsb
Because you guys don't confirm you tried 2T I mention this again.

When I tried to run 400 FSB in sync(lol almost typed in 1:1 from my DDR2 days) with both my E3110 and also my Qx9650, I left the memory timings on auto and it autoed to 9-10-10-28 2T as the SPD on my RAM goes only to 763(DDR3 1526). But still it was a no go. But I guess in my case it was different coz, the RAM was at fault here.

[webwilli]

Dang.. aint the sky falling down on ASUS today!
I got around the same barrier on my E8400...

is it a CPU or Boardproblem?? or must i just find the right setting in bios?

[TheHardCase]

Asus bios 507 really works, lol.
Viper Patriot 4 x 1GB Sticks, PVS32G1600LLK
Stock settings 1600MHz 7-7-7-18 @ 2T 1.8 VDimm Don't even think about running it there.

OverClock to 1800MHz 7-6-5-18 @ 1T 2.2 VDimm & wow, there is alot more.

[eternal_fantasy]

Asus bios 507 really works, lol.
Viper Patriot 4 x 1GB Sticks, PVS32G1600LLK
Stock settings 1600MHz 7-7-7-18 @ 2T 1.8 VDimm Don't even think about running it there.

OverClock to 1800MHz 7-6-5-18 @ 1T 2.2 VDimm & wow, there is alot more.

Can you post your BIOS settings for us to try out?
Thanks!

[saaya]

Reading through the ASUS forum S2E thread:
http://vip.asus.com/forum/view.aspx?...e=en-us&page=3
Looks like the data corruption when the FSB is raised from stock 333Mhz is very widespread on the ASUS Striker II Extreme... and I'm not seeing ghosts and tooth fairies benching/tweaking early into the morning!!

did they try ide drives?
im pretty sure its a sata issue, nvidia always had sata issues since nf3 iirc, with nf4 they were quite bad, especially with raid, with nf5 and nf6 they seemed fixed and only minor problems appeared, not with 7 it seems to be broken again... lets hope it can be fixed with a new bios...

And Saaya, I see it important (vcore, vdroop) mostly for us that do OC their cpus. I own an striker extreme 680i and sometimes you just hate that your CPU is stable at 1.36v but because it goes down to 1.34, you cannot keep it stable and that means by the time you get it stable, it will be getting a lot more energy than the neccesary when idle.

afaik it doesnt, it gets a slightly higher voltage, but the amps are very low in idle. its really just a psychological thing... afaik running a cpu under load with 1.4v is WAY WAY more stressfull for the cpu than running idle at 1.5v and low temps...

100% sure
i benched high FSB on both SATA and IDE drive
no difference
only difference is higher PCI Express frequency on IDE hence why i like using it more

so then pciE is still linked to SATA hmmmm
well its strange people are getting this much coruption.
yes, its normal to corrupt an os when running unstable mem speeds, but mind you, ive been booting and ocing in windows with setfsb to unstable speeds runnng orthos for weeks, and had the system freeze or reboot at least 200 times a day, and orthos or setfsb corrupted every 2 days or so, but i didnt have a corrupt windows xp install even after all this for 3 weeks or so...
so there is definately a higher corruption rate here than usual, even if you push the speeds too high and mem isnt stable.

i corrupted the drives on really tight timings past 2GHz on ref board and i know that it is RAM related 100%

the people that have coruption issues here are running in the 1600s to 1800s with cas7... from what ive seen they know what speed their mem is stable at in other boards, and id be surprised if its stable at lower speeds in 790...
i dont think its a mem problem, remember for fubarswe , afaik, the system was rockstable and passed stability tests for hours, but then eventually the system would slow down and then get corrupted. and the corruption vanished when he switched to an ide drive...
im pretty sure its a chipset thing...
it might be memory too, but... that wouldnt explain why that many people in this short time are having so severe corruption issues. if the memory is that unstable it causes corruption your usually way unstable and cant run any 3dmarks or other serious benchmarks besides superpi and everest.
if you push your system clocks up only benching everest or superpi, and then you blow your install, you wouldnt even post about it here would you?
its to be expected if you do that...
to me all those corruption issues happened unexpected, people thougth the system was stable but then suddenly the install went belly up...

i have the same problem now.
i only change the cpu today:
old: E6600, max FSB on STRIIEX = 490 FSB, 400 FSB rockstable all the time
new: E8500, more than 375 = no post
i don´t understand

hmmmm ask OBR for his settings, he got 45nm quads to 490fsb
what bios are you using and what settings?

Early reviews over the web show the 790i on par with X48, in some cases ahead, others just behind.

well, and x48 = x38, and x38 = p35 minus some pciE lanes so...
790 = roughly as fast as p35/x38/x48
from what ive seen 790 can reach slightly higher mem clocks with cas8 (~10%) and slightly higher fsb speeds (~10%) has slightly better mem performance at 1600 and 1800 memory speeds (~10%) but in all real world benchmarks and apps ive seen it surprisingly doesnt stick out compared to p35/x38/x48. for SLI 790 is the way to go, if you run singlecard or xfire it doesnt really matter if you go 790 or intel, the performance will be about the same.

to sum it up:
790 is interesting for people who want to see huge ddr3 clocks with cas8, people who want to run sli, people who want to get higher fsbs, and people who like nvidia chipsets/boards overall and prefer them over intel based boards.
now keep in mind this is all based on stuff i read and feedback i saw from a few dozen people and webistes, i havent actually played with a 790 board myself yet. i merely summed up what other people told me or posted on forums or reviews and websites about 790

Asus bios 507 really works, lol.
Viper Patriot 4 x 1GB Sticks, PVS32G1600LLK
Stock settings 1600MHz 7-7-7-18 @ 2T 1.8 VDimm Don't even think about running it there.
OverClock to 1800MHz 7-6-5-18 @ 1T 2.2 VDimm & wow, there is alot more.

very nice! i see you managed to get the bandwidth up!

[Leeghoofd]

NuclearJock posted the 0509 bios for you lads, no changelog yet :

http://www.xtremesystems.org/forums/...d.php?t=182965

[cognoscenti]

LMAO! 8-12-12? does it need trcd and tras that high? :o

was posting here for advice

[eternal_fantasy]

did they try ide drives?
im pretty sure its a sata issue, nvidia always had sata issues since nf3 iirc, with nf4 they were quite bad, especially with raid, with nf5 and nf6 they seemed fixed and only minor problems appeared, not with 7 it seems to be broken again... lets hope it can be fixed with a new bios...

I myself Use a dedicates PCIe RAID controller, so don't think at least in my case linked to SATA issues. Note that most if not all of the corruption happened on the ASUS S2E, after the FSB is overclocked, even though the Memory is at relaxed timings.

the people that have coruption issues here are running in the 1600s to 1800s with cas7... from what ive seen they know what speed their mem is stable at in other boards, and id be surprised if its stable at lower speeds in 790...

Again using relaxed memory speed and timings, the S2E continues to corrupt system data as soon as the FSB is raised. CellShock-1800@1600-8-7-6-21-1.9V should not cause corruption. Reference boards have no such problem as far as I can see.

People continue to report high overclocks on their ASUS S2E. I can do exactly the same thing, posting hours of ORTHOS runs with Screenshot of CPU-Z showing 4.05Ghz CPU, 450FSB, 1800-7-7-6 Memory. Followed by a . And maybe even a . BUT all of those are mute points when data are corrupted by the motherboard.


A screenshot of a scan for system file integrity should also accompany screenshot of your system settings after at least a few hours since the overclock settings has been applied, after some system restarts and stability test runs.

To run system file integrity scan:
- search for "cmd"
- right click-run as administrator
- type in "sfc /scannow"
after scanning it will tell you if your windows is corrupted and weather it was able to fix the corrupted data(most probably not).

[webwilli]

hmmmm ask OBR for his settings, he got 45nm quads to 490fsb
what bios are you using and what settings?

i hope OBR read my post and give me some help

joe_cool from awardfabrik have over 500 FSB with the E8500: http://www.awardfabrik.de/forum/show...8&postcount=47

i ask him too for the settings.

yesterday i first test default clock: 3.166 with 333 FSB
then i raise the fsb to 401 FSB and WTF: no boot.
first i think the cpu was bad and will not run with 3.800 GHz. so i decrease the multi to 6 and still no boot with 401.

i not test much settings. it was too late for me and i go to sleep.

[noobzed]

Is nvidia planning to give official P03R2 (P04?) anyone knows ?

edit: Just see this on Everest Ultimate..

But set to 1,95V Dimm in bios, and bios say it is 1,95 too
So wonder if Everest right or wrong..
SPP temp is strange too 118°c..
Should I drop voltage or not ?
My Corsair 1800C7 is rated for 2,0V

[KTE]

Hmm.. data corruption, I hate that. No way I would risk it yet until something is sorted out or more clear about this. MEM instability can easily screw up a Windows install though, and so can unstable FSBs. I've had it on many boards, not just the 790i you're seeing it now, I experienced it on my P35 about 6 weeks ago, "system32 is corrupt or missing" on reboot after trying 475FSB with an E8400.

cognoscenti: he's kidding but trying to say something at the same time
Have you tried 8-8-8 MEM settings?

[saaya]

anyone know how to force flash the 790i evga with the p03r2 bios?

tony posted the commands necessary earlier in this thread, page 15 or so

was posting here for advice

sorry man, i was just joking, no offense or anything
i had to laugh when i saw those timings since they are so ridiculously high

I myself Use a dedicates PCIe RAID controller, so don't think at least in my case linked to SATA issues. Note that most if not all of the corruption happened on the ASUS S2E, after the FSB is overclocked, even though the Memory is at relaxed timings.

hmmmm i see... so maybe the board is tweaked to agressively, so it barely is stable at stock but as soon as you oc some timings are too agressive and cause problems. or its a divider issue... did you compare mem linked unlinked and checked if it makes a difference?

noobzed, if you want to know what vdimm your board is giving use a multimeter to meassure it. dont ever trust the onboard monitor in bios or windows or everest. they are good ans easy to access, but not 100% reliable. they all ready their values from the same chip on the board, and depending on how good its implemented and how well the bios interprets the values it can be spot on or completely off.

[Broly]

saaya i tried that but uponr estart after flash its still same bios. awdflash gives me the commands as if i typed help when i put in /F/PROT am I using the wrong version? i am using the exe that came with p03 bios iso

[eternal_fantasy]

hmmmm i see... so maybe the board is tweaked to agressively, so it barely is stable at stock but as soon as you oc some timings are too agressive and cause problems. or its a divider issue... did you compare mem linked unlinked and checked if it makes a difference?

I think your first point makes sense. I have used linked and unlinked memory dividers (mostly linked&synced) but both causes corruption... Hopefully a solid BIOS release will sort these problems out

[cognoscenti]

sorry man, i was just joking, no offense or anything
i had to laugh when i saw those timings since they are so ridiculously high

Well I am not a man but no harm done

This board seems really nice so far but when it comes to ram I'm still very much a learner.

[slim142]

So guys I have good news. My Q9450 RETAIL just arrived!!! Im really happy!!!

Lets see what the box says

FPO/Batch#: L803A777
VERSION#: E34934-001 PACK DATE: 03/14/08

Hmmm, what batchs did you guys get for your Q9450 (retail)? Are the current batchs OCing good?
Also, I have noticed it says 1.25v Max, I thought it was 1.35v? I must have gotten confused with a Kentsfield...

[saaya]

saaya i tried that but uponr estart after flash its still same bios. awdflash gives me the commands as if i typed help when i put in /F/PROT am I using the wrong version? i am using the exe that came with p03 bios iso

sorry dude, i guess you need a special exe to flash...
i dont have it and dont know anybody who does

I think your first point makes sense. I have used linked and unlinked memory dividers (mostly linked&synced) but both causes corruption... Hopefully a solid BIOS release will sort these problems out

well nvidia somehow seems to have managed to fix this as on reference boards there is no corruption right?

seems like shamino is also having corruption problems with 790...

Well I am not a man but no harm done

This board seems really nice so far but when it comes to ram I'm still very much a learner.

oh, your a dudette then?
memory tweaking really easy to learn, its quite easy to just go by trial and error with memtest i wouldnt recommend starting to tweak ram in windows, if you dont know what your doing youll blow your install and its really frustrating to recover or reinstall

What about DRAM drive strength ?
Could that have an influence on the RAM degradation/failure ?

im very sure it does... basically drive strengths changes on how fast the voltage is pulled up and down.

if think of the traces inside a chip as channels, and the signals going through it are water waves, then there is a spot in the channel that is elevated so the water cant pass it. you need to get some water over it so a signal is recognized.

the higher you elevate the water level the easier it is to get a wave to dip over the elevation, but you also increase the power the waves hit the elevation with and you increase the wear off on the walls of the water channels. so increasing vdimm helps to get a signal through, but it can wear off and damage the traces or chip.

id say changing the drive strength is like changing the shape of the wave.
using the same amount of water and using the same level of water you can get the waves over the elevation in the channel and can get the signal through. but at the same time a steeper wave most likely wears off the walls of the channel and the elevated spot more than a wave using the same mass of water but having a lower thicker shape.

still id say with propper drive strength tweaking it should be able to reach the same signal quality as a more massive wave with a higher water level but has a lower chance of damaging or wearing of the channel walls or the elevated spots. but hey, this is pure speculation, i really dont know sht about chip design and mfging proces.

i know that some electronics are way more sensitive and get damaged rather by quick steep voltage spikes than lower and more steadily increasing voltage waves...

THC, does the ocz mem work in another board?
check the ocz support forum or pm ryder.
if they get hotter using the same voltage, and they use the same chips, then it sounds like they are damaged using high vdimm. thats the only thing i know that can make one stick run hotter than another one with the same chips... electromigration...

[BenchZowner]

Plunge taken....Striker II Extreme ordered... hopefully it'll make it through my review tests alive

[Broly]

Anyone know what version of awdflash accepts the /prot command?

[cptkoulis]

currently "playing" with my 8500 on S2E using 9,5x multiplier

when setting FSB to 425 in NVCP,system froze several times

w/o changing anything else,i set cpu@1800(aka 450FSB) and system boots/works fine

[GAR]

Does anyone recommend running 4x1GB of OCZ 1600 memory, at 450-475fsb, thats where i will be with my Q9450.