I.nfraR.ed

Here's the video Thought somebody will ask, so made one long time ago. The only thing that changes is the specification field. Name stays the same - XP-M. Haven't tried with older version though. When I bench it next time will try.

I had a video somewhere. Board has soft L12/self-modded bios + hard L12 and mobile mods via switches. When I flip the mobile mod switch board detects it as XP-M. Doesn't happen with other Thoroughbreds. The chips is exactly the one shown in the picture. Uploading video now...

https://www.dropbox.com/sh/1ct53ethtbz3g63/RfpixGyvrk - link to photos I've used 100kΩ VR for both mods. Nothing hard in terms of modding - just FB to GND. First photo shows vcore mod, second one - vNB (black wire is GND, white on the IC is the FB and the other white is measurement point for NB. Last photo shows measurement point for vcore. Be aware that I have blown one of the MOSFETs - it might not be related to the mod, but it also might be. Got replacements, so I will fix the board soon. I think they overheated during the 5600+ session due to the fact I forgot to put a fan over the heatsink and it (the heatsink) was slightly bent to one of the sides. Hate those plastic pushpins Gigabyte uses, but didn't replace them with bolts like I do on other boards.

Very strong CPU, can't beat that frequency.

You didn't say anything about GPUTool Community Technology Preview 1. It is my tool of choice for HD48xx series. Did it have the same limit?

Problem is I don't have much free time recently. Don't know if I will have time to rerun.

sorry guys, but I'm not in Bulgaria right now. I have 2 mods on the board - vcore and nb mod. Will do vdimm mod when I get back home. Can show you pics next week.

I have hardware mod. Board limits k8 to 1.55.

Great cpu! Don't think I can beat that with my crappy eff with my chip

It might seems so, but... what is the difference between 1220 and 1220 SE? Is it only the TDP/default voltage that differentiate them? 1220 release date is February 2007, while 1220 SE (both F2 and F3) had been released on August 15 2006. Question is - could cpuz distinguish between both models?

Try gputool from techpowerup.

CB at around -48 on the evap.

None. It's the only ClawHammer I have. The other one (FX-55) died a long time ago. Still have some Winchesters, but all are low-end models - 3000+ and 3200+. Edit: Nah, there are people like Knut and Mafio who have better CPUs and I'm no near a master... This score is crap, should be sub 25s. BTW, I'm waiting for 2 FX-55 from good steppings, hopefully 3400-3500+ capable on SS.

Rotting Christ - Iwa Voodoo (Κατά τον δαίμονα εαυτού) Seems my Clawhammer likes it, too

Thanks. Not that efficient on an old xp install, but I will rerun later with the other cpu. This one has better IMC and CPU-NB, but weaker cores.

Don't get me wrong, I really appreciate the hard work you did, but is there any specific reason you didn't do a full-pot LN2 test instead of -150, thus eliminating the need of constant pouring while trying to keep temperature at -150? Plus full pot will be more interesting since Ivy goes all the way down and beyond LN2 temps, not mentioning all AMD chips since Deneb.

Too bad I sold my crappy 8150

No review, just a personal testing in order to decide which thermal compound I should stick to. Unfortunately I was only measuring the pot base temperature with the probe I usually use. No IHS readings, but I used the maximum clocks as a measure, don't even remember what was the bench. Can't provide any real data Tried GC-Extreme and Ceramique 2 with Deneb at full pot as well and didn't see any difference there, too.

Strange, but according to my testing GC-Extreme and Ceramique 2 were on par with 3770K at -186 and 6+ GHz, while the Noctua seemed to be completely frozen and the clocks were pretty bad. And I tested all of them twice each.

IMC has a bad coldbug, yes. Had to put a heatsink as a buffer and control the temperature with the tension of the bracket .

https://dl.dropbox.com/u/30576042/PST_3_4_1.ZIP

With Thunderbird?

Damn, I was so close with SS, had even 54.74x but couldn't save it. Anyway, soon LN2

Do you really read what he wrote? When you tie a record, you're both the record-holder, whether it was on the same race or after 10 years. What you say is simply not true. Both athletes/racers get the recognition. PS: It was addressed to PAQd.

Yes, same for me. Had another issue too - when opening several instances on a Thunderbid CPU at high FSB (250+) it just randomly closes, while an older version works like a charm. Seemed like stability issue at first, but it wasn't.