I.nfraR.ed

Yes, my fault. But I've been running several times this way...you know, it happens once...

Using THE tweak

I think it's dead now. Had a leak around the chipset and now it doesn't post. Let it stay overnight with removed battery. All voltages are still in tact, but no POST

281 is what it did before, but I'll try to beat it, even though it would be hard.

Don't remember how good is my AN7, might check it. I think it was able to load Windows somewhere between 255 and 260.

It can run SuperPi at 269 - 270 Greatest of all s.A Abit boards I've ever had. Trying to keep it alive as long as possible.

I use kneaded eraser and once bought a board that had nasty vaseline on it. Spent so much time removing it, so the eraser could finally seal on the pcb And...btw, pictures are only visible to members of the oc'ers forum.

Nice, you put Supreme-HF on the cpu

I don't call him cheater in any means. I don't even know if the score is ok or not, it's beyond my knowledge. I just never messed with those settings and didn't know it's not visible if you use non-default set of settings. There are many similar scores with similar clock in the top, so if he's a cheater than all of them are, lol I think fps in subtests are ok for the given clocks, compared to the other guys at the top.

Damn, I've wondered why I'm not able to match some scores, even with higher clocks. 3DMark06 is pretty consistent across different runs, unless you get a bugged run.

What I did wrong? I think you're being a little bit harsh here, there are many noobs at hwbot (including me), not only pro guys. And still, we are living under the same sun... This format with the efficiency thing is not applicable, for the reasons massman wrote. The best approach is like it is currently done. Overclocking is for all things together, but the main thing is pushing everything to its maximum. Cooling separation can't be done as well, people can give a picture of watercooled system, but in fact could have used SS.

Thanks! In fact, only one important thing is missing - info about SLI (enabled/disabled), but hope it's clear from the other 2 submissions (03 & 06) that only one card is used. I don't have other 05 screens, it was a single run. And I have only one GTX280.

Ok, I have the same question and don't want to open a new thread. Was running 06, 03 and 05 with the GTX280 and it seems I've accidentally switched to sensor tab on the 05 submission - noticed when submitting Is it a problem? Frequencies are visible, the VGA model is visible as well and there's only one variant of GTX280. The submission in question: http://hwbot.org/submission/2261164_i.nfrar.ed_3dmark05_geforce_gtx_280_41644_marks Here's the 06, same card used, same clocks: http://hwbot.org/submission/2261163_

That's not leveling it up, either. It's just who's fastest and who has a lot of free time. The other approach is better - you have a reasonal cap for cpu frequency and memory speed/timings, then you fight for efficiency.

Ok, thanks for the info. Seems I don't know much about later Abit stuff But still own the NF7-S that did 281 FSB

Done! Talking about minimum purchase it might be hard for me to sign that. If for example I'm using 45L a month now (mostly because of time constraints and lack of own transportation) and I'm required to use 100L/month, it just can't happen if I need to use the same method of delivery - taking a taxi to the depot and then go back. I'm on a regular full-day job and I can't afford too much absence. Currently taxi costs me more than the price of these 15L LN2 I'm able to get (borrowed old dewar from a friend), but if they offer me a cheap dewar rental plus delivery to home address, I'd use much more LN2. Imagine FX-8150 at 6 cores, 7GHz and I have only 15L...

Why not think from another point of view... HWBOT competitions usually are motivating me to bench some specific hardware, that I could postpone forever. I don't do it for the prizes, I do it for pure fun or the national pride in case of CountryCup. Plus there are always plenty of interesting discussions going on. Another good thing is I can always learn something new or improve my skills. These are more than enough reasons for me to participate in competitions. Oh, and it's always pleasure to poke/joke HondaCity, Massman or Christian Ney, haha.

I don't have a Gigabyte board as well. Couldn't find a cheap decent used board, so I'm buying tomorrow 970A-UD3 from the store. That's not the best board you can use, but I will still participate just for fun. Hope I won't be too frustrated with it. Haven't used Gigabyte since 790FXA-UD5P (this is actually the only one board I've ever had from that vendor, lol). I don't think everybody have 990FXA-UD7 or similar high-end boards... Not very faimiliar with current GB lineup, but 970 chipset should be ok for this kind of competition (no multi-card setups required).

Sure I'm a PCMark05 noob .

That could be better. However I don't know if AM2 boards have a chance - not played much with Brisbane/Windsor. Ask Knut about this . If you make it so, I'd have 2 stages to compete in, otherwise it would be s.A only Or could search to buy AM2 board.

There are no AM2+/AM3 boards At least AM2 boards' cpu support list stops at X2 models (Brisbane/Windsor). On the other hand, s.939 is not included .

Haven't seen anyone raised that question. I wonder why there are global points for 5x, 6x, 8x, 12x and so on number of cores? Hardware points - ok, but as I've read somewhere PCMark05 utilizes up to 4 cores. It's not like wprime. Correct me if I'm wrong here. So, a guy with "1 of 1" score in unpopular category still gets full global points. He can even run at stock, the outcome would be the same. On the other hand if you manage to score 4th in a tough category like wprime1024 with a single core processor, you get less points. How fair is that? Well, if there are not global points for these kind of systems, they would be not appealing for submissions, that's also true. Not going against anyone personally. If there are flaws, they will be always exploited. Just sharing my thoughts. I'm not even a points collector, I just do it mostly for fun. In my opinion there should be global points for PCMark05 as a whole, but not based on core count.

Here it is http://hwbot.org/submission/2260449_i.nfrar.ed_pifast_athlon_64_4200_x2_manchester_35.83_sec All these 2D scores with that cpu were single sucessful runs, I'm sure they can be improved further, especially wprime. Haha, yes . Soon you'll have to buy other models as well, lol. You can see my temp at the first photo from submissions with the Manchester. Need to bench properly all these s.939 cpus in the box. Some of them seem very decent. Most of them are single cores though.

Oh, I disassembled everything yesterday, because I need to focus on the Gigabyte challenge. But I can use water cooling anyday LN2 is a little different story - I pay more for the taxi then the actual LN2 cost . And I have only 15L dewar, that is borrowed from a friend. But...if your challenge is interesting, I would accept it

Yeah, It doesn't work for me with 1090T as well. I've never ran PCMark05 seriously, but wanted to try those transparent windows tweaks. Well, I tried it on my 24/7 system, so that might be the problem, but I see gains from mouse moving only.