S_A_V

Good work

What frequency you see on the box of the CPU? What frequency you see in CPUZ (Specification field)? What frequency CPU working when all cores/threads 100% loaded (on motherboards that not support fixed CPU multi for Turbo Boost)? The stock rating, not the turbo rating. At least four: 1. Stock speed 2. Turbo with only 1 core enabled 3. Turbo with only 2 cores enabled 4. Turbo with 3 or more cores enabled

Yes, you can. Post scores under "2x 5870 CF" category and don't forget to pull down list of cards in GPUZ on screenshots.

Nice score for only 1150 GPU clock

Pieter-Jan, what BIOS settings do you use for 2000+ MHz memory clock on Asus C4F? My board can do only 1920 stable and 1959 for cpuz valid (dual channel, unganged mode): http://valid.canardpc.com/show_oc.php?id=1209633 Tried 1:4, 3:10, 3:8 memory dividers - all ~1920 stable, ~1950 can load OS but unstable, 2000 - no post In single channnel with one memory stick can do 2000 only for cpuz valid. Memory tested on X58/P55 before and can overclock up to 2050 6-7-6. BIOS 0801. Memory in black slots. May be memory overclock limited by IMC of my 965 C3? Will try on 1090T X6 later... EDIT: yes, looks like Thuban IMC slightly better than Deneb C3... 2006 6-7-6-18 1T with 1090T X6: http://valid.canardpc.com/show_oc.php?id=1210637

Yes, it's special BIOS for GOOC2010 No, it's Q3QP (3.06 GHz) chip. All Q3FE binned at 2.4 GHz. No, it's ES.

Thanks! I used BIOS version 1.7, boot at 361 MHz HTT and then "OC Dial" button on motherboard up to 415 MHz.

Multimeter. Ok, I made this pic for you: Pencil R710 resistor for raise Vmem. Measure Vmem on one of two capacitors marked "Vmem measure points"

Thanks. I flashed XFX card with Asus BIOS and then Asus Smart Doctor used for gpu voltage. Memory voltage raised with pencil mod.

GPUZ still show native BIOS version and DeviceID of this 9800GTX+

Every time you pressing PrintScreen key after aquamark run, screenshot saves not only in system clipboard, but also as .tga-file in this folder: %USERPROFILE%\My Documents\My Pictures\AquaMark3\ (Windows XP, Vista) or \Users\<user_name>\Pictures\AquaMark3\ (Win7) So, if your system frozen after you hit PrintScreen, but before you save detailed screenshot from mspaint, you can reboot and restore original aquamark screenshot from .tga-file. Try search for *.tga files in "Documents and Settings" folder (Windows XP, Vista) or "Users" folder (Win7).

bump 2 weeks passed but mipmapped results are not removed yet

When my score was reported I added link to archive with aquamark files: http://rapidshare.com/files/284040370/am3_single_radeon_hd4650_ddr3.rar.html 242005 on Windows XP and 253049 with Vista (with affinity tweak). Both scores done at the same clocks, both not bugged and repeatable. The only reason of high GPU score is vista. Also I have 251351 in vista score with 40 MHz less clock on CPU.

Very good score for cheap cards on AMD setup, nice job! What temps and voltages on GPUs?

Vommaro-PC used mipmap with all his results on 8500GT 512Mb DDR2. There is no way to bump fps in nature of 3dmark2001 by 30% higher than others in top5 with less video memory clock. These cards are limited by memory clock mainly, not CPU or even GPU clock.

980 on 295 say enough for me. Voted for Youngpro.

Shortcut to my profile still on front page so everything is fine Didn't add any scores in few last weeks, my DFI is dead and I still not found any replacement for it

Checksum is correct. Checksum is incorrect Checksum checker - http://www.techpowerup.com/superpi/

"In total, 469 results have changed in rank or points, causing you to lose 739.5 hwboints in the past 7 days." lol it looks like new point distribution system (look at multi-gpu categories) or bug.

Game Tests GT1 - Wings of Fury 773.18 FPS GT2 - Battle of Proxycon 812.54 FPS GT3 - Troll's Lair 312.35 FPS GT4 - Mother Nature 631.42 FPS GT2 and GT4 too high for single card, GT3 too low for SLI. Bugged score?

Yeah, they do real damage It was my Palit 6800GS based on the same PCB as any other Palit/Gainward 7800GT/6800GS including Gainward Bliss 7800GT 512Mb GLH. Vgpu was not so big - around 1.7V. PMW just can't handle current and burn.

Anyway Gainward Bliss model not so much better then reference 7800GT cards. Better memory ICs but worse GPU PMW. Try to set 2.0V or higher Vgpu on Palit/Gainward 7800GT/6800GS and you will see flying mosfets with firework special effects Reference 7800GT cards can be benched with 2.0V Vgpu and will survive if cooling is good enough. That is why we got beaten only in 3dmark01 and AM3 but not in 3dmark03/05.

Only Gainward Bliss 7800GT 512Mb Goes Like Hell (Part number = NE/7800TXT352-PM8070-GLH) with 1.4-ns memory clocked to 1300MHz by default: http://www.3dnews.ru/video/gainward_7800gt_512/print/ Palit 7800GT 512Mb with 1000MHz default memory clock not better than reference 256Mb cards: http://www.nix.ru/autocatalog/svga_nvidia/PCIE_DDR_GeForce_7800GT_DualDVI_SLI_47480.html

Reference 7800GTX 256MB - 1.6-ns mem. Reference 7800GTX 512MB - 1.1-ns mem. No. I think videocards can be separated by memory type (ddr-2-3-4-5) or even bus width (64-128-256), but not by memory size. Especially ones with more than 256Mb! Why they must be separated if they performs the same in benchmarks? Just because we (I have results on 7800GT 256Mb too) got beaten by the owner of better card (Gainward Bliss)?

How many vendors released 7800GT 512MB? Gainward, Palit, CardExpert, etc.? They are all the same. Different brands, same manufacturer. 7800GTX 256MB and 7800GTX 512MB is another story. They was totally different - PCB design, memory, cooling. 7800GTX 512MB officially exist as nvidia reference card. 7800GT 512MB was initiative from Gainward. Everyone can make mistakes. Even hardware moderators at hwbot. 3dmarks and aquamark3 not use more than 256Mb video memory.