Memory not fully detected

That's fine, I don't really understand it either :D but there is an obvious difference that I've attempted to explain in the txt file.

You can remove that email ad now by the way, file has been sent.
 
Well, I can't make out a lot. But you're right that there's defintley a hint in there somewhere.

What differnce do the tests make when they are run in batches of 8 vs batches of 2? As in do the batches of 8 test something more and push something more than the batches of 2 do?

Sorry, a bit new to prime95 :)

Playing Rise of legends right now. I think thread optimization may have made a difference. Couple of times its crashed early, but right now again, it seems to be lasting, for a lil while at least.

The artifcats seem to be related to the drivers. Gotten pretty bad with this update. Even the load screen and main menu is showing them now. I may be reverting to earlier ones soon.
 
Last edited:
I no wiser than yourself on the matter, I just thought it a significant and interesting discrepancy, I'd imagine the batches of 8 stress the system a little more, but don't know for sure.

As I said earlier, all I really know about prime is that its useful for quickly isolating weaknesses in the system, especially with the memory and CPU.

Oh, I also know that it attemps to seach out prime numbers and that FFT
stands for fast fourier transform, whatever that is :p .
 
Just thought i'd give a shot at runnign Sp2004 as well since In oticed that's what you were using.

Just a question, do you run two copies with each setting at "any CPU" ordo you specify 0 for one and 1 for the other?
 
Just about to stop the test manually at 5 hours.

One of them is running fine, batches of 10 (priority was 1, I wasn't aware this made a difference. Is this what determines the no. of batches?)
The other, going very very slowly. Onlky gone through the first 4 Tests. Barely using any system memory. only about 10mb. Other one is taking about 950 mb
not sure if this is how its supposed to be or?

This is without running 3dmark. Just on it's own.
this stresses only cpu and ram, right? yea nm, stupid question. Just checked the sitenote given along with blend on sp2004 :p

does 3dmark stress only gpu? if so, what are your results when you run only that? (sorry if you already said this earlier)
 
Last edited:
Yes, that's correct.

The priority option determines how intensely the the system is dedicated to running the task, the higher the number, the more resources it comsumes. I have it on 8 as this is the highest it can be without slowing the system to a crawl.

If you're running two copies at once, best to run 1 blend test and 1 small/large fft test, this way you ensure both cores are being fully utilized all the time.

I'm trying this right now, just to make absolutely sure that all is well with my CPU/RAM :confused: :) .

I'd imagine you're running 2 blend tests at the moment, which is why one is running very slowly as even with two gigs of RAM, it's having to sling data in and out of the swap file on your hard disk, when this happens, one core is often not made full use of.

I don't know what would happen if I just ran 3dmark without prime, my guess is that it would not lock up, I'll give this a try later.

3dmark stresses the whole system, but mainly the GPUs, in much the same way that Prime stresses other things such as PCI bus etc, but mainly the CPU/memory.
 
Last edited:
balth_dire said:
Yes, that's correct.

The priority option determines how intensely the the system is dedicated to running the task, the higher the number, the more resources it comsumes. I have it on 8 as this is the highest it can be without slowing the system to a crawl.

If you're running two copies at once, best to run 1 blend test and 1 small/large fft test, this way you ensure both cores are being fully utilized all the time.

I'm trying this right now, just to make absolutely sure that all is well with my CPU/RAM :confused: :) .

I'd imagine you're running 2 blend tests at the moment, which is why one is running very slowly as even with two gigs of RAM, it's having to sling data in and out of the swap file on your hard disk, when this happens, one core is often not made full use of.


I don't know what would happen if I just ran 3dmark without prime, my guess is that it would not lock up.
Click to expand...

If that's the case, that indicates that everything is fine with the hardware and it's all down to compatibility. Seems to me like that BIOS updates are our only hope of fixing this. I doubt nvidia drivers would help any (maybe my legends artifacting) since I don't think they would deal with mobo compatibility.

We should take Quad SLI users and march down the Asus offices :p
 
I think so. We've certainly gone as far as we can.

I've found at least five other people with the same board who are trying to run QUAD, which must be five more than are working in Asus's technical department :p . I'd say we have the advantage!

What happened to that Asus man by the way :D ?
 
lol. sad, but very possibly true! We should just join the Asus support team :p We've probably narrowed it down more than them. I hope they aren't thinking "it's all nvidia's fault. we'l just wait for them to release better drivers"

Good question about that asus guy, lol. That shop I took it to never called me back. I couldn't be bothered chasing it up because I figured that with your help, i've gone way past anything he could do. I called him once to tell him that the 305 beta BIOS fixes the ram problem, but he was unavilable and then supposed to call me back, which he didn't.

If i do call him up now, he'l just say, i bet, "yea, i talked to him. he doesn't have any idea. sorry"
 
Hello there.

Just a quick update.

The other day, I decided to do a clean install of Windows XP x64 as I noticed that Quad SLI drivers have finally been made available for the OS.

In brief, I used:

Nforce Chipset drivers 6.86 (the latest release for xp x64, downloaded directly from Nvidia's website)

and

Forceware 91.92 (not on Nvidia's website but available by way of "a certain guru of the 3rd dimension" ;) )

So far with this combination, I've yet to experience any of the problems we've been having with xp x32, no Oblivion crashes as yet and no 3dmark lockups.

Whether it's because Nvidia has done something fundamentally different with these drivers, or Quad still isn't fully enabled on xp x64 - even though it appears to be - I've yet to ascertain, but things do seem significantly more stable.

You might want to "acquire" a copy of xp x64 and see how you fare. I believe most games are compatible.
 
Last edited:
Hmm, I might try that out. Would it still be fine if I just installed it alongside this one? Or do you need a clean format for this solution to work you think? I know you can dual boot easily and all, just wondering if format is needed for this particular solution.

Format might be better, but i've made loads of small changes now to this, and have it customized etc etc, too much of a pain to go through it again :p

Thanks for the update though, it'l take me a few days to try it out, but i'l do it soon as I can.

I did update my chipset drivers, but I think the crash still happened. (again, with Rise of Legends, but it's very possible it's just a game thing)
I've been running Oblivion without any lockups at all lately.
 
I'm pretty certain now that the crashes are different manifestations of the same fault. When autosave (on travel) is enabled in Oblivion, it's making some very specific demand on the system, which, when Quad is also enabled is causing the lockup/crash to occur.

Whatever this demand is must also be being made in my 3dmark/Prime test and by certain other games - Rise of Legends being one - again though, because the crashes are so sporadic, it's virtually impossible to identify precisely what the "demand" is.

All I can say is that neither issue (3dmark/oblivion) has thus far occured in xp x64, probably a coincidence :rolleyes:, but it would have to be a pretty hefty one.

Yes, you can install xp x64 along side xp (that's what I've done), just make sure you choose a seperate partition and that memory remap is enabled in the bios. :)
 
hey, i'm yet to try the x64, i'l probably manage to by monday. Just anotehr update though. Today, my comp's hung three times, all three times in windows, which is very strange. One common theme running, was that I was playing a video. (maybe just coincedence, dunno)

Only possible thing I can think of, is those 91.92 drivers I installed earlier cause haven't really changed much else. I'l change back like I was planning to, and see if the problem disappears.

Glad to know that you're problem is fixed though! Thanks for letting me know :)

edit - you sure we should be getting 6.86? those are the drivers for Nforce4 Intel x16. Shouldn't we be getting Nforce4 SLI Intel? I think that's what CPU-Z says the chipset is.
 
Last edited:
Definitely the x16 drivers, these apply to Nforce 4 boards with the dual full length PCI-E x16 slots like ours and the A8N32-SLI. Cpu-z is a good, but ultimately universal tool, hence the occasional lack of accuracy! :)

As far as solving things, don't speak too soon:

A: It may be an extreme case of luck.

and

B: The drivers for xp x64 might not have Quad properly implemented yet. The green bars and logo are showing up and the results I obtained in 3dmark were all but identical in each OS, but I can only test upto 1600x1200 at present, not the optimal resolution for QUAD. I'm keen to know how you get on at 1900x**** and above ;) .
 
Hehe, okies :) Nice to have it confirmed, I don't really remember why but I do remember that when I updated my chipset drivers, it was to 6.86.

Just carry out a test for me if you can please, take a video on your comp (preferablly non DVD), and just play it for a few hours at a strech? Doesn't have to be full screen or anything, but just playing. That's one of the common things that was going on earlier when my comp. hung in windows and just possibly even earlier when it happened.
 
OK,

Do you know what the format of the video you were playing was? Mpg, avi (divx or xvid)? Also what player were you using? Media Player, VLC, Realplayer? let's try and be as consistent as possible.
 
It was avi in one case and I was using windows media player classic (ver. 6.4.9.0) In the three other cases, once was during youtube I think, and twice was while streaming from Shoutcast TV on winamp.
Thanks a lot :)
 
Right.

Well I'm a Youtube obsessive and have been using that a fair bit over the past few days, no crashes thus far.

I did hear that having certain audio codecs installed can make things a little twitchy, can't recall if this was specifically with Nforce boards though.

I've got a fair few Divx's I was planning on viewing, so I'll give one of those a go.
 
Hello again. ;)

Apologies for the extremely belated response - unexpected amount of work!

I tried playing two full length films (Xvid) and didn't get a crash, nor have I yet had a crash/reset on Youtube. Of course, this probably means, once again, that I've just been lucky rather than the problem not exsisting with my setup.

I notice that there has been yet another driver released, 95.XX or something, though after reading the official list of changes - which includes three irrelevant and frankly, utterly useless modifications - I didn't bother installing.

Let me know how you're getting on when you can.

Have you bothered with xp x64 yet?

Also, were you planning on trying one of these new nf590 or 680 based boards? The 590 chipset is supposed to have overclocking issues and it looks as if only Asus are going ahead with their release. The 680, (which "replaces" it) is supposed to materialise at the beginning of November, hmmmm ;) :rolleyes:.

Best wishes.
 
You must log in or register to reply here.
Back
Top Bottom