poparamiro

max 1560mhz 3DM2003 on ~ 18 oC water, do not have CPU for 3Dmark 2003 with ~ 0 oC water i think 1600mhz possible, low volt.

Those 3 phase need good cooling and can sustain 1.5v+ an air/water, no nned for zombie (if the card VRM is OK, of course). Even two phase VRM with very good cooling can sustain a G92 GPU with big air voltages: Check below: http://hwbot.org/submission/2345165_poparamiro_3dmark_vantage___performance_geforce_8800_gt_256_mb_8975_marks If you have e golden board with defective VRM, is wothed to zombified.

I know vga's like my own pocket, i know what and how they scale, i know voltages and everything about them. I can not be fooled to believe (only with very solid proofs) that with normal water the board goes 980mhz and with water at 4 oC 1110mhz. This is not possible. With 4870 and 4850 is hard to do 1000mhz even on cold water (with best of them, binned), 1110 is just out of range. By the way, what was the voltage at 1110mhz?

Thanks

I have no problem with 4 oC chiller, i have no secret, just water cooled to above 0 oC by ambient. Just rerun and show everything like in my video (especialy GPU freq and temp on AFB monitoring). It not have to be exactly the same freq 4 oC water means about 20 oC max on load, at ~ 1.5v I hope you are not upset by my request, but 1110 mhz with 4850 at ~ 20 oC gpu temp is unbilievable. I was ask to film a lot o my results, and of course i do it, but no one was that incredible as your 1110mhz 4850. Best regards

Probably uP controler, find pins with resistive divisor (to gnd and v out) with values from 500 ohm to 2-3k, most likely ~ 1k and mod the pins to 20x to gnd. What card is that?

There you go: enthusiast league with 1100+mhz HD4850 on water http://hwbot.org/submission/2522209_flashg_3dmark_vantage___performance_radeon_hd_4850_gddr3_15051_marks

Sorry, no disrespect about your work, but no 4850 on the planet will do 1100mhz on water. Film the run with AFB monitoring, clear show the sistem, GPU temp and frequency like in the vid below:

Found this, was done back in the time that i filmed everything, a lot of reports about air water frequencies

No, just the memory freq count big on 3850 and 3870. Temp bug on that card, ~ 12-15c load i think (~ 2-3c water)

Thanks For me yes, i do not like software vmods, or extra wires and software (hotwire). I like simple and quick control of VGA.

Yes, same on matrix, good luck.

It is the Matrix? You mod the vCPU from PWM (pin 4) ? http://hwbot.org/image/1144388.jpg

Pump the voltage and change the cooling Lab501 forum for help (what is your user?)

GPU-Z bug, there are two cards, check picture. Best regards

My vote go to disolvation of the enthusiast league, to much controverse. I will continue "clocking", no problem.

GT2 takes 1200+mb of vram, 1gb or less cards have not enough vram. I have also terrible framerates on 1gb cards (try 560, 5850 ..etc) Some cards with ~ 1200mb vram begin with good frames and at half test or so the slideshow begin =) Example: sub 2fps here, normal for category 3/3+ fps Same situation for sli (ex 2x 5850) very low frames. Did not investigate much.

==)) I pointed to GT3, not very hard to see the enormous bug

848 fps GT3 - very nice

72% asic

Was just for testing mods, aircooling.

I tested tonight some mods for Asus 290X DC2 Gain 110mhz on Elpida with 1.7v, more do not help. GPU tested to 1.45v load on air

Some problems with drivers. All ok now Gain 110mhz on Elpida - 1.7v, going up no scaling.

Thanks Tested to 1.75v, no improvements. With good GPU i think 1330+ mhz possible on water.

Asus DC2