could you run that IBT with half the clusters disabled
every other core at those clocks.
Printable View
Blacklash: The 6100 should produce less heat, you may be able to get 4.6 on aftermarket air (maybe better:)).
redrumy3: Sure give me a little time and I'll post them up, ofc your system will vary, you'll probably still have to tweak a little.
Manicdan: Thats a good idea, Thanks! I'm a little burned out after yesterday but I'll be sure and try that... :up:
demonkevy666: I'll try that too! :yepp: I saw some core options in bios but I didn't really look closely yet.
Hey dave nice stuff yesterday.. I was up all night Friday and like you I needed a break. Im on again trying again with my rig.
first things first. On friday my 1st mistake was grabbing this cheap rosewill cooler and hastily strapping it on. I wasnt planning on using air too long but when I ran 1.4vcore and my temps were getting to 67F I knew something wasn't right. Then today I remembered a few years back I got this massive cooler from scythe to review. do you guys remember the OROCHI?
Dude I found it in the box in the kitchen cabinet and low and behold the amd plate was already mounted. Check this out.
I went in the bios and kept the same clocks and temps after the massive air cooler. I was getting 51c before and now i have like 31c. I would love to join you guys on H20 but I dont plan on keeping this on too long this was meant for Ln2. but until I get this figured out Its nice to have some better air cooling. Right now my main goal is trying to get this thing up to 5ghz and doing some wprime runs. right now its at 4.8. the guide has been an asset thus far. :up:
before
http://i239.photobucket.com/albums/f...h/ROSEWILL.jpg
after
http://i239.photobucket.com/albums/f...ith/orocHI.jpg
Well, it looks like the screen cap feature in bios isn't working in 0810, and I don't have a good enough camera for pics... :(
Basically, just turn off turbo, and all the power saving features (cNq etc).
Heres a pic showing most of the voltages and settings. Hope this helps.
In the end your going to have to tweak your own gear anyway, no 2 systems are the same. ;)
http://i30.photobucket.com/albums/c3...20Settings.png
what cooler for my setup?
1)Xigmatek DK+Ultra Kaze 3000rpm (now at my 1090T, there I have no problem with temps under full load)
2) venemousX+Ultra Kaze 3000RPM
3)Corsair H100+2x Ultra Kaze 3000rpm
whoa sin you did ln2 last night? yeah man send me some footage to work with. I am actually eager to compare my sony vegas 10 rendering times to my xeon 3540. I dont have any of that installed on the BD rig just stripped xp and basic necessities. I cant wait to see your ln2 session. Im benching pc05 and pc vantage right now off a single c300 and a gtx 460 at stock clocks atm just so I have a foundation. Once that is done I might have to get my mega raid rig on here. (4x crucial m4 raid0 on lsi 9260-4i w FP key.) I must say I think what helped me the most from the guide was disabling core boost and the mem clock. That had me scratching my ass for hours on friday. I would look at my ram clocks in cpuz when I set 7-7-7- and they were showing at 2-14-13- in cpuz
@flanker just grab the heaviest one that has the heaviest base and find the amd clips for. put your highest fans on it and thats it. just grab heaviest one. next open the windows and put a warm sweater on and crank some AC/DC like our AU team friends do. Im listening to dirty deeds and for those about to rock record. I prefer Bon Scott.
Hmm I have 0810 and screen capture works for me. Are you trying to save to a flash drive or? It saves my screen caps as .BMP files to my flash drive.
Basically we are trying to see all settings you have modified to come to your clocks such as Loadline options.
Any help is greatly appreciated.
Has anyone else had a problem with their FX CPU still throttling down after overclocking on a Gigabyte board? I could swear I read about someone else who noticed this and found a way to defeat it. As it stands, no matter what I overclock to, cores on my 8120 will flicker back to a 14x multiplier when under load. There's no telling how much of the potential performance that's sapping.
Interesting. Mine seems to regardless of those settings. C1E, C6, CnQ, CPB, and HTC are all off but it drops down to 14x under load at times. The higher I set the CPU frequency, the more cores are throttled and the longer they spend there. When everything is set to 3.1, it never seems to happen or happens briefly enough that I don't catch it on CoreTemp. The same is true at 3.2. At 3.3 it starts happening on one core here or there. By 3.6 it happens on as many as four cores at a time for brief periods. By 4.3 GHz, most cores are throttled down and most of the time. After about 4.3 GHz, the effect starts penalizing me more than the increased frequency helps. CB drops from 6.2 to 5.9 for example.
It's like the TDP estimator part of Turbo is still on even when I turn turbo off.
mine throttled with an earlier BIOS the F5 BIOS, with F6E it didn't throttle anymore.
Trans am, is that the PIS 2200 Cas 7-10-10-28 1.65v T2 kit? I used that same kit! hahaha
yea I did some LN2 last night, ill get my results up on here and some pics, gotta let that videos upload. Ill have it up later tonight or tomorrow. i made sort of like a LN2 prep guide for those who are new and want to try it out. it is something i could have really used when i went to try it.
Aha, that explains it, sin0822. The UD3 has F5B as the latest. I wonder if there's a beta available?
from what I saw there were some Betas up, and then they were taken down. I am not quite sure about your board tho, but I am sure some news BETAs or finals will be up. This is a pretty popular platform, so BIOS support should be constant I would think.
demonkevy666: Can't disable single cores in bios... Only disable options are 3+4, 5+6 and 7+8, so basically I can only shut down whole modules. :(
Seems if the Flash drive is formatted as NTFS it's not recognised.... :p:
Heres the important stuff:
http://i30.photobucket.com/albums/c3...1017004243.jpg
http://i30.photobucket.com/albums/c3...1017004306.jpg
http://i30.photobucket.com/albums/c3...1017004401.jpg
Do people find a big difference in clocking between the 8150 and 8120?
Thanks for nice information. I have a unsolved question though :p: (regarding this image)
Attachment 121329
Right after see this table, I guess there's some bug in a chip regarding some operation clock speed which sticks together with HTT, L2 speed for instance.
So I've tried the same test with a hope of making another excuse for Bulldozer's low IPC at default HTT.
My comparison settings are:
1. 3.6GHz (200x18) / NB/HT 2.4GHz (200x12) / DRAM 800MHz (200x4)
2. 3.6GHz (240x15) / NB/HT 2.4GHz (240x10) / DRAM 800MHz (240x3.33)
3. 3.6GHz (266x13.5) / NB/HT 2.4GHz (266x9) / DRAM 888MHz (266x3.33)
4. 3.6GHz (266x13.5) / NB/HT 2.4GHz (266x9) / DRAM 712MHz (266x2.66)
Note that for 3 & 4, dram speeds don't stick exactly with 800MHz, so I tested twice for FSB266 under two different dram speeds both distant samely from around 800MHz.
Anyway, my results are totally indifferent - means that higher FSB doesn't make sense at all :(
(I'm so disappointed that I cannot even make a graph......)
I tested Cache & Memory benchmark, CPU Queen, CPU AES, FPU Julia, FPU Mandel in AIDA64 and Cinebench R10 / R11.5.
Do you think you have any other variable which I (probably) missed? - I've only controlled CPU multiplier, NB/HT multipler and DRAM multiplier along with HTT clock.
GA-990FXA-UD7 - F6
GA-990FXA-UD5 - F6
GA-990FXA-D3 - F4
GA-990FXA-UD3 - F5
GA-990FXA-UD3(v.1.2) - FB
GA-990XA-UD3 - F9
GA-990A-UD3 - F4
GA-990A-D3 - F7
-Update AGESA 1.1.0.0 code
-14.Oct 11
FX is no more NB 3000+ (like X6)
for normal Air/Water cooling?
From what I've seen they OC very similar atm. The 8120 even seem to do slightly better over all..
Still too few samples to draw any real conclusion. But 4.6-4.7 on air seem to be the norm. Higher is hit or miss at this stage.
A list with model + OC and default VID would be cool.
It could be board variance.
I noticed before that jacking ht on some boards and or fsb impacts diff boards. It has to do with ram straps and how diff manufacturers adress ram differently.
I'ts rather simple, when you jack up the fsb and use a lower divider and mainly auto subtimings, some boards just when strap is applied force lower subtimings.
Also if he is using what appears to be B2 F die and you are using retail G die that can cause some conflicting data ;)