Ya I actually did a quick test for scaling. See the following links:Quote:
Originally Posted by chew*
1x 4870
2x 4870
3x 4870
Printable View
Thank you ben, perfect example of what I was trying to explain. Happen to have a vantage comparison? Also quick question ben, I know 3d06 isn't linear with 3 cards but how are actual applic ations/games? Is it the same trend in your personal experience?Quote:
Originally Posted by Ben_Brown
I only have screen shots for 1x 4870 and 3x 4870. However when I ran 2x 4870 it was like high 13's to low 14's if I remember correctly.Quote:
Originally Posted by chew*
Thx, looks to be more linear, 2nd card 4k 3rd card another 4k.
I'm not sure you understand how the SPD is supposed to work.Quote:
Originally Posted by Aquineas
Maybe I don't. Care to enlighten me? Here's what I bought.. So maybe it can run at the higher speeds, but it still bugs me that every mb (all three of which are rated at DDR2 1066) report it as DDR2 800.Quote:
Originally Posted by MrMojoZ
your running too hot, and the chip could be failing stress at oc settings as a result, in my opinion. please try to get this temperature down and then try to re-oc.Quote:
Originally Posted by Cybercat
take an average of your temp using ITE smartguardian and CoreTemp. It would also be a good idea to get a thermal probe on the heatsink/waterblock if you are able to, so you can add in to the calculated average.
Had some fun last night in my below freezing garage :D
While this is not prime stable by any means, I am none the less happy that I was able to complete this along with half a dozen or so SuperPI 1m runs before it crashed. Especially with this (ES) chip.
Model - 940 (ES)
Stepping - unknown but best guess is a C1
Frequency - 4309MHz
vcore - 1.712
NB Frequency - FACOC AC 0825FFM
RAM -2 x 2GB
MB - Asus M3A79-T
Cooling - Water + garage in sub zero weather
Windows - W7 x86
Okay I will try but this is in non-techie laymans terms so please don't flame me ..........Quote:
Originally Posted by Aquineas
DDR2 JEDEC spec only goes up to DD2-800 PC2-6400. So when set to Auto memory settings all MB's will set RAM to 800Mhz. If you look at the SPD tab in CPUz you will see that the profile for 533 is called EPP as it is not JEDEC approved. By setting your BIOS to Manual you can up the RAM to 533 to get DDR2-1066 / PC2-8500 (but also noted that you may need more vDimm too).
Wrong, jedec spec goes up to 1066 for ddr II, Tony from OCZ helped get this passed...Quote:
Originally Posted by Russ_64
Aquineas, Tony from OCZ is a very very helpfull person on these forums, I'm sure if you contacted him or post in the OCZ forum here he would be more than happy to "help" flash your ram so it's detected @ 1066 by SPD.Quote:
Originally Posted by Aquineas
Ben awsome job man, I know you must have froze your butt off, I know i did last night, beat my int el chip by 80mhz to with the same cooling ( frigid air ) and did a 32m, all i could do was 1m ;) Thats what I like to see.
just took out the Phenom 940, stuck back in the 9850.....hmmm spending 375.00(CAD) for New CPU, for this much difference, with same board.....lol. Mind you I did get ask tho if I did want to sell my CPU, hard to find around here....
Wprime with PII 940
http://img233.imageshack.us/img233/5465/wprime1ew0.png
Wprime with 9850
http://img233.imageshack.us/img233/5...ime9850vv8.png
Hondaguy, hehe I have seen the same difference.. which is just that the phenom2 overclocks better than phenom1s. :(
Altought my wprime result with 24/7 clocks is this: http://faior.miumau.net/wprime3.6.JPG
and 24/7 screen: http://faior.miumau.net/3.666.jpg edit url
Cybercat might I make one other suggestion, Your using 8 gigs with all 4 dimm slots populated, this tends to add stress to the mem0ry controller.( yes i know how to spell )
Try running only 4 gigs and see what you can acchieve, I know you probably have no plans to run 24/7 with 4 gigs but it would help determine what is limitng your OC.
There is a BIOS update for anyone running a
M3A78-T BIOS 0802
01. Support ACC for AM2+ 45nm CPU
02. Fix the problem that the system may show boot fail message if ACC enabled.
03. Fix the problem that AM2 CPU Vcore voltage may be incorrect if change CPU voltage in AI Suite.
04. Improve the system performance when use certain CPU.
05. Support new CPUs. Please refer to our website at: http://support.asus.com/cpusupport/cpusupport.aspx
(ACC is a overclock technique which can adjust the voltage and frequency for AMD CPU directly under OS)
interesting. all who have this board, please report back on the effects of external ACC. will be very interesting to see!Quote:
Originally Posted by HondaGuy
Dude, THANKS VERY MUCH for your post! ASUS have been on a roll with new Bios support. When you said the M3A78-T had a new bios, I checked for my "non -T" M3A78. I thought I had the newest Bios for my board. The one I have is dated Dec. 8 2008. They released a new version on the 31st (0901)that I didn't know about. I'm not sure what it added, but I am going to install it as soon as I finish this post.Quote:
Originally Posted by HondaGuy
@Chew ....I ordered my Phenom II 940 today. I can take my curse off all of you now.:D......
http://img183.imageshack.us/img183/8...7346um8.th.jpg
My RAM still hasn't shown up though. I ordered 4GB of OCZ Gold series from newegg with Egg Saver shipping last Sunday and still haven't got it yet.:mad: I will never use their EggSaver shipping again. :down:
What board are you using, I guess the M3A79-T? Not alot of difference there in Wprime still have here the SB600 board also running PII on,, thinking of selling chip thoQuote:
Originally Posted by faior
Guys, I think Ive once again some retarded problem.
Like last night, I was testing 3.8Ghz CPU, NB 2Ghz (since 1.8Ghz would limit the CPU to 3.6Ghz) and HT at 1.8Ghz. RAM at DDR800, so pretty much only multiplier OCíng.
I set Intelburntest for 5 runs, 4 runs were all showing stability and then in between the 4th and 5th run I got a BSOD, 0x00000124. Ive gotten this one quite often, I googled it, only resulting in more questionmarks as it seems this BSOD could be for about anything.
A very few seemed to be related to the soundcard, most others were to certain series of nVidia cards, unstable OC's etc. But sometimes I ran a succesfull 8 run Intelburntest and then the very same settings BSOD'd the 2nd time after the 2nd run:confused:
Today I decided to get rid of my soundcard and uninstalled drivers. Then I tested my RAM, only for 40% coverage though, about 1.5hours, without a single error. I just run Prime95 for 1.5hours without issues. Although I dont know how Intelburntest relates to Prime95, this is quite weird nonetheless.
I also read a few single OC results from Intel users that adding FSB VTT and NB Volts (NB as in motherboard of course) helped them getting rid of this BSOD rage.
Thus far it seems to be all fine and happy but Ive this curse everytime Im convinced it's fixed it happens some way again. Does anyone have experience with this particular BSOD? Any tips or whatever?
Im even thinking that Intelburntest might not be such a good program after all. It's nice, but Prime95 managed to get my load temps higher and I cant believe that 15 minutes of Intelburntest defeats say a 8hr run of Prime95:shrug:
Anyway, having no sound at this point, thinking Ill use my onboard sound after. Saves me from a load of crap anyway and onboard sound ain't all that bad nowadays:rolleyes:
Hmmm ramm, vid card related in some cases, sounds like PCI or PCI traffic.
Curious if the PCI lock works, is it set to auto? sometimes at auto or default 100 it's not really locked, ruuning it manual at 101 can sometimes lock it.
I do get that exact code when really pushing this rig, more voltage does not fix it for me, bringing NB temps lower does....
Doesn't seem like SB, usually the easiest way to determing SB issues is your hard drive will "katunk" and BSOD.
Ain't getting BSOD's anymore now, my PC restarted this test I just started instead of BSODíng.
Ive PCI-e at 100, but Ill set it to 101 then just to be sure. GPU related, highly unlikely to be honest. There could have died a million things but GPU, no:p:
Anyway, seems like I hit my max clocks for CPU then with +0.150V, 3.7Ghz (or slightly above that but it just restarted at 3.755:rolleyes:).
[QUOTE=Rammsteiner;3582887]
Ive PCI-e at 100, but Ill set it to 101 then just to be sure. GPU related, highly unlikely to be honest. There could have died a million things but GPU, no:p:
Can be gpu related if the GPU can not handle a high PCI frequency ;) not all gpu related errors are due to actual ocing of the card.
Ben nice cold run :D
honda 400 mhz is much of a change
I went from 2.7+ish to 3.7+ so I had almost a 1ghz clock speed change so it was a needed upgrade...
Update for Oldguy932 :)
http://img88.imageshack.us/img88/207...681hrr2.th.jpg
http://img88.imageshack.us/img88/478...h21cpu2pz8.png
:rolleyes: Phenom II 940 BE 0850BPEW @3768MHz/1.46V NB2368MHz/1.175V
Vista Business x64
AIR TT120 Ultra +12cm FAN
room temp ~20C
PS. on the temp graph you can see when my Central Heating stops working :p:
i have constant restarts, it rarely just gives an error or something.Quote:
Originally Posted by Rammsteiner
and the max temp for the cpu goes down with higher frequency.
insane cooling is required to give the cpu a working temp of 40 degree or something.
for me 3700 and has a crash temp of about 44 degrees but 3500 doesnt even crash at 51 degree.
i guess i still do something wrong, as lightman has 51 degree on 3750 :(
Well often it fails before it even gets a chance to heat up. Like if I try 3.5 at stock vcore, I'm lucky if the OS boots. I've been using smartguardian to monitor temps. Core Temp would act funky (temp would get stuck it seems). Anyway, other than the floor fan I don't have many options to lower temps, but I'll post back here if I think of something. I don't own an external temp probe.Quote:
Originally Posted by pershoot
Yeah I've played around the with memory. Underclocked it, took some sticks out, etc. I didn't try memtest like you suggested earlier though. Either way it didn't make any difference in stability. It seemed like a memory problem when it kept BSOD (memory dump) but never failed prime. But I don't think it is. I appreciate your help, I think either there are some BIOS issues or I just have a dud. Anandtech was able to get 3.9GHz with stock cooling on my board with an ES. :shakes:Quote:
Originally Posted by chew*