I.nfraR.ed

Some of the utilities are useless. I have some utils even more useless. Others are "must have" for old platforms and some I haven't even used, only collected through the years. Thanks for the descriptions, I had no time for that, posted the links just before I left town.

Dropbox Download (click link) Athlon64 Tweaker nForce2 Tweaker (by CodeRed) Ati HD38xx Bios Edit BiosMod (GF2 bios editor) Central Brain Identifier Clockgen NF2 Clockgen NF3 Clockgen NF4 Clockgen 1.0.5.3 CPU Tweaker Dothan Clockgen ERAMN220 (the xp/2000 ramdisk Japanese oc'ers used in the past/not sure about now, extract with e.g. 7Zip) GeForce123Edit (it was for GF and GF2 series I believe) + GeForce123 Runtime nHancer 2.2.1 nTimings 1.0 NVTweak 1.7.1 Powerstrip RaBiT 2.2.1 Riva TNT Clock TweakBios 153b Voltage Factory by AwardFabrik VoltageTune 1.1.2.1 vt1165 (plugin for RivaTuner, HD4890 uses this voltage regulators) X-BIOS Editor 1.0 RC3 b520 ZTAGP Tool (tool for AGP in chipsets like BX440, KT133, AMD 760 etc.) AMD CPUID AMD CPU Info CPU Heat CPUMSR CrystalCPUID RMClock v17 WCPUID Hot CPU Tester Coolbits.reg (unlock overclock in Forceware drivers) Should have much more somewhere... Also have many s.A mod bioses for several popular boards. Also the bioses for DFI NF4 boards lineup.

Congrats to the winners and all participants. It was very hard in terms of money and effort. Had a lot of problems in almost all stages. Shame that Belgium is infront of us. Alby, will beat you next year! Special thanks to Wiz for the tips, never refused to answer when I asked something . And big thanks to the guys (Alkoholok, Rager, Grigor) who helped with hardware. Even if they were not the greatest samples, we still managed to close almost all stages.

Nice setup, but bad luck. Next year will be better. Still a nice score with this monster cpu!

Great one, Vivi!

Unreachable for me. Great one!

Nah, it's not nice. Need better board for bclk overclocking, this one is not suitable

Yeah, same monster chip, too bad we couldn't push it to the max. What you see was just a random clock as a start and the board died after the 3rd run. Great score, Wiz .

990FX-UD3 rev.3 had some problem with "high"(1.2V+) volts on the CPU-NB - it was simply shutting down. And yes, it was throttling at high load, I checked with Cinebench. 970A-UD3 rev.1 has enormous vdroop and 990FXA-UD3 rev.4 (the prize from the Giga comp) is horrible, but haven't played enough with it to tell what exactly is wrong. Bios seems worse than the one on UD3. I wrote Gigabyte, but no response. I've only tested it on water, but what the old Crosshair IV Formula can do, the Gigabyte board can't even touch. I suspect it's the bios, because there's no more overheating, the cooling is improved versus v.3 and I guess VRMs are different. It might be me doing something wrong, but there was not enough time for pretest. In the end the little 970A-UD3 was the best choice from all 3, but due to the big vdroop (0.07V) and not so good FX-8120, the scores are mediocre. Also the CPU-NB frequency is tied to the HTT frequency for whatever reason (can't be higher), but it might just be the latest beta bios, have to check again F7. 990FXA-UD3 rev.3 was able to pull 200 to 300MHz more in multithread with 1090T in the past, but it seems damaged now.

We were benching last 2 or 3 consecutive nights without getting any sleep (work during the day) and when the stages were finally over, I went to bed in the evening and woke up at 6 next morning...12+ hours of sleep I was hoping we can get that APU stage, but Indonesia managed to beat 3 of our scores. We could do better, but the asrock board died just 15 minutes after the start with the first APU to bench (6400K). It was really a gem with the best IMC from all our APUs (1400+) and really strong GPU (1400MHz 1.35V on water). 1800MHz should have been possible on LN2. Next day we got the Asus board, but there was not enough time to rebench it, if it's still alive. Other scores could have been better too, but we were limited in time and resources. For Aquamark stage we didn't have a good Bulldozer and Gigabyte boards were throttling with really high clocks, so I'm happy we managed to get the 3rd place. For 3DMark we simply didn't have enough time and suitable hardware. I tried to get some good score with the HD3850 on LN2, but something went wrong, so I had to post the single score on air produced 30 minutes before that. Not sure, but maybe the only team that managed to clock the memory on these stupid PCI 610 cards GunGod managed to get the Llano Catzilla score literally 10 minutes before the end. Additional 300-400 points were possible there as well. Thanks to my teammates for the hard work, now need to concentrate on the other stages

We have problems submitting Aquamark score, cpu is @ 3809.13mhz but when upload it data base read it @3809130 MHz screenshot is ok all is ok but i cant submit it say "smthng got wrong" any clues? 10x

I always have a fan over it and it was fine in the previous competition, but it failed pretty quick in this one. Just like the Gigabyte board long time ago...

VRMs didn't get warm at all and our A85 Extreme6 just died out of the blue - no condensation, the board was even insulated on the bottom side with liquid tape. cpu volts were 1.6V, vGPU 1.45V and it died 15 minutes after the start. No particular reason, just tried to change a setting in bios, then it started showing weird post codes on the display and after the third restart never posted again.

Thanks for the honest answer. Your performance is good. Roman's chip throttled I think, but haven't compared loop by loop. I had the same with my 8040 1M run. I can probably do better, especially in multi-thread benchmarks, but need CH5F. Gigabyte boards I have surrender much earlier.

I have one as well, but my favorite s.A board is NF-7, so this one is just staying in the box, waiting for its turn. Push it more, 265 should be possible out of the box.

How many did you bin?

6800K. No way to do the same with the smaller models. Not reachable even on LN2 I guess?

I got 4K on water with cpu at 4.95 and gpu/mem at 1260/1300.

It is on the passive stock heatsink (yes, the same Zotac card, no fan attached, no mods), but it took several days to find how to overclock ram and it went through all team members Driver might be important, or I'm just trolling

Kenny, I was able to submit 2.26 score yesterday. Did it change today? We have some scores that are not submitted yet.

trololo pretest, max on stock cooler

Well, ok, but 400+ MHz difference? I saved the log, but the system is disassembled now. Will tell you what is written there later.

How is it unstable when with the old wrapper I can run 7.5 and the new wrapper complains about it even at 7GHz? Same settings in both cases. With several tries sometimes it runs, but it's 1:10 or so. When really high freq it does not run at all.

At cpu clock > 7-7.1GHz the new wrapper keeps telling me to restore original benchmark files when I try to run the bench. Reinstalling doesn't fix it. Log does not say anything useful. With 2.26 7.5GHz is ok, please don't discontinue 2.26

He validated it on the 3rd CU (CU2). Nonetheless amazing cpu.