|
|
|
Video Card Thread (Mostly Gecube x1950xt)
|
|
|
MaccerM
Member
|
16. August 2007 @ 02:54 |
Link to this message
|
|
Well my Freezer 7 and HR-03 arrived yesterday and got the fitted - what a difference! My cpu was idling around 48c with stock intel cooler, now its 49-50 under load o/c'd to 2.02ghz (with stock AC goo too)! The best bit is the HR-03 though. Full load temps at 710/760 are 56-57c core 39-40c on the PCB. With the peltier heap that was on there I was getting 75c+ core and 54-56c on the PCB, and that was at 675/702!! All that extra heat from the TEC was trying to kill my CPU too.
HR-03 - I cannot sing it's praises enough!
3Dmark is now 5581 and ATitool reckons there's 740 in the chip, and over 800 in the vram. I'm quite confident now that my E6320 will hit the 2.3ish the m/b FSB will allow (as I've got DDR2 667 with a 1066 chip it should really be all about the cpu & m/b)
Best c£sh I've ever spent on cooling!
Q9550 @ 3.4Ghz-1.275v & TRUE, Asus P5N-T Deluxe, 4Gb Dominator 1066 (5-5-5-15), Inno3d OC 8800GTs in SLi with TT Duorbs, 580w 5.1 Setup, 2x DVD-RAM, 320&160Gb SATA2 HDDs, 17" 3ms Hyundai TFT, 700w Jeantech PSU.
|
|
Advertisement
|
 |
|
|
AfterDawn Addict
4 product reviews
|
16. August 2007 @ 03:01 |
Link to this message
|
|
Isn't it awesome? If a year ago, somebody told me you could get away with cooling a 130W Graphics card with your case fans at 500rpm (that's half their minimum speed) and no fan on the graphics card either, I would have laughed in their face.
The Freezer 7 Pro is also an excellent cooler, but on the strength of the HR-03, I'm tempted to try the Ultra 120, now they're dipping in price.
|
Senior Member
|
16. August 2007 @ 08:23 |
Link to this message
|
|
well, I tried O'cing my card up as much as it could go manually with ati tool, and I could get to 950, and then i would get artifacts and my computer locked up and crashed...
Still dont know how people get to 1500-2000mhz though...
As for the Hr-03, thats some pretty good cooling man! Do you have any fans on it or is it just a heatsink?
|
AfterDawn Addict
4 product reviews
|
16. August 2007 @ 08:25 |
Link to this message
|
|
Ahem, 950 in Atitool is the base frequency of the memory. That's actually 1900mhz.
Damn, my card only goes up to 1746!
|
Senior Member
|
16. August 2007 @ 09:25 |
Link to this message
|
woah woah woah... Are you telling me I can get to 1900mhz??!!!
Thats awsome!!
I still dont get what you mean by "base frequency"
Quote:
Ahem, 950 in Atitool is the base frequency of the memory.
|
AfterDawn Addict
4 product reviews
|
20. August 2007 @ 03:14 |
Link to this message
|
|
All I meant was, that's the frequency of the memory bus. Since the memory is DDR (Double data rate), the memory itself runs at double that speed, i.e. 1900mhz.
|
|
MaccerM
Member
|
20. August 2007 @ 03:43 |
Link to this message
|
|
As Sam says RAM is DDR nowadays, so the base frequency for say an effective 1.4ghz is actually 700mhz. I.e. you double your real to get your effective clock freqency which accounts for the double data rate effect, so yeah you're at 1.9ghz! - pretty sweet!
With my HR-03 I have a 92mm fan on there. Noise takes a back seat to cooling in my system priorities but it's still quieter than the pelt!
I think I'll stick with the freezer for the CPU. If it'll keep my cpu under 55c when I get up near the top overclock (which is only 340 vs 266) and I'm pretty sure it will.
Another interesting point re the power consumption of the X1950XT agp peltier version. My new psu has a usage display on the back and with the pelt on there it used about 240-260w under load. With the pelt off and my new cooling in it tops out at 185-190w! I think that's absolutely staggaring, not only does that old pelt try and set fire to your other components as well as your gfx card, it uses an extra 70w in the process!
Q9550 @ 3.4Ghz-1.275v & TRUE, Asus P5N-T Deluxe, 4Gb Dominator 1066 (5-5-5-15), Inno3d OC 8800GTs in SLi with TT Duorbs, 580w 5.1 Setup, 2x DVD-RAM, 320&160Gb SATA2 HDDs, 17" 3ms Hyundai TFT, 700w Jeantech PSU.
|
AfterDawn Addict
4 product reviews
|
20. August 2007 @ 03:46 |
Link to this message
|
|
Which is why peltiers shouldn't be used for PC cooling, they're horribly inefficient!
Where do you think that extra 70W goes? Heat!
Did you realise that you can get away with no fan at all on the HR-03? I do, and it runs fine!
|
|
MaccerM
Member
|
20. August 2007 @ 03:56 |
Link to this message
|
|
Yeah that pelt has got to be the worst cooler I have ever seen. I had read that the TEC used 20w or so before so I was really surprised at a 70w difference! It stands to reason though that the best way to cool something down is not going to include something which makes more heat!
Not like the HR-03 though! I'm sure you can run fanless fine but as I say noise doesn't bother me and I'd rather know that my temps were gonna stay low, (especially since the card I have has already taken more heat punishment than it deserves!) What temps do you get in your sys?
Q9550 @ 3.4Ghz-1.275v & TRUE, Asus P5N-T Deluxe, 4Gb Dominator 1066 (5-5-5-15), Inno3d OC 8800GTs in SLi with TT Duorbs, 580w 5.1 Setup, 2x DVD-RAM, 320&160Gb SATA2 HDDs, 17" 3ms Hyundai TFT, 700w Jeantech PSU.
|
AfterDawn Addict
4 product reviews
|
20. August 2007 @ 04:20 |
Link to this message
|
|
With my case fans at 500rpm (lower than any silent fan you can normally buy), at idle around 50-55C, Full load around 70-75C, a good 15C lower than with the stock heatpipe cooler.
|
|
MaccerM
Member
|
20. August 2007 @ 04:49 |
Link to this message
|
|
Yeah, mid 70s which is about what I got with the pelt cooler after the tape mods. For me I'd rather have the fan on with my HR-03. It's fairly quiet but mainly I never go above 60c at full load and I have the extra o/c headroom too. Still it is impressive that a passive heatsink bests the stock cooler by that much, with only a few mice blowing air very gently into your case!
Q9550 @ 3.4Ghz-1.275v & TRUE, Asus P5N-T Deluxe, 4Gb Dominator 1066 (5-5-5-15), Inno3d OC 8800GTs in SLi with TT Duorbs, 580w 5.1 Setup, 2x DVD-RAM, 320&160Gb SATA2 HDDs, 17" 3ms Hyundai TFT, 700w Jeantech PSU.
|
AfterDawn Addict
4 product reviews
|
20. August 2007 @ 05:37 |
Link to this message
|
|
True, but there's no extra overclocking headroom to be made form the better temps, they're exactly the same as with the stock cooler, I think the limit is more due to the components themselves rather than the cooling, but in any case it's nice to have an almost silent graphics card now (I say almost since the memory chips buzz slightly during games, a lot of X1900 owners have reported that)
|
|
MaccerM
Member
|
20. August 2007 @ 06:12 |
Link to this message
|
|
I suppose that's right, the lower temps just allow me to actually push the components now. I couldn't even run 675mhz when I first got the card because the heat issue was so bad, so I maybe don't have any extra headroom but I can at least access it now!
Q9550 @ 3.4Ghz-1.275v & TRUE, Asus P5N-T Deluxe, 4Gb Dominator 1066 (5-5-5-15), Inno3d OC 8800GTs in SLi with TT Duorbs, 580w 5.1 Setup, 2x DVD-RAM, 320&160Gb SATA2 HDDs, 17" 3ms Hyundai TFT, 700w Jeantech PSU.
|
Senior Member
|
20. August 2007 @ 06:19 |
Link to this message
|
Well I got bored, so I decided to post some benchmarks that I had gotten with the TideWater on full rpms. The FPS were taken from ATITool's "show 3d view" and the Temps are from RivaTuner. (The case temps are incorrect, I got those from RivaTuner too, but my case was allways at 86F max, or 30C.)
Overclocking
Core: 702Mhz 635AVG FPS @ 62.75C Core
Mem: 837Mhz 46C Case
42.75C Idle
------------------------------------------
Core: 702Mhz 701AVG FPS @ 62.75C Core
Mem: 927Mhz 45C Case
44.25C Idle
------------------------------------------
Core: 702Mhz 667AVG FPS @ 62.75C Core
Mem: 855Mhz 45C Case
42.63 Idle
------------------------------------------
All overclocking was also done with ATiTool.
I dont play with my memory at 927Mhz, I get artifacts there, but i think my highest was 855, that was the most stable. I think sammorris said that was equal to 1900Mhz, because 855 is the "base".
Can you explain that sammorris?
|
Senior Member
|
20. August 2007 @ 17:10 |
Link to this message
|
|
***To Everyone Who Is Having Crappy/Undesired Performance W/ Their x1950xt!***
I have the solution! (Oh man, have I been waiting for something to show how the x1950xt can perform when all of its power is harnessed!)
Ok, heres what I did...
-- Get the newest Omega Drivers from www.omegadrivers.net
-- Run driver/install
-- After fails, right click on desktop, and select Properties
-- Go to Settings, then Advanced
-- Click on Adapter tab, you should see your Adapter type (e.g. RADEON X1950XT Series)
-- Click on Properties button
-- Click on Driver tab
-- Click on Update Driver... button
-- Click on "No, not this time" button (to bypass Hardware Wizard searching the web)
-- Click Next
-- Click on "Install from a list or specific location (Advanced)" button
-- Click Next
-- Click on "Don't search. I will choose the driver to install" button
-- Click Next
-- Click on Have Disk...
-- Click Browse...
-- Navigate to the location where ATI installed the driver files. For example, for me they reside in C:\Program Files\Radeon Omega Drivers\v3.8.360\2KXP_INF\CX2K_RAD_OMEGA
-- Select the file.
-- Click Open
-- Click Ok
(The next part is a little confusing, but seemed to work for me)
A dialog comes up with a huge list of drivers. I Navigated until I found the RADEON X1950XT.
-- Click Next
At this point a driver warning dialog popped up that recommended against installing the driver because Windows could not verify it, but I threw caution to the wind and installed it anyway
-- Click Finish
* * * Whew * * * that was painful.
Then, let Windows reboot your pc, and when it starts up again, log on. It may take up to 30 seconds for the driver to finish "refreshing", as the screen may turn on and off a couple of times, just wait, that is normal.
Basically, thats all I did to get my x1950xt working like it should be.
I went from playing GRAW 2 at 15fps on low @800x600 to 25-30fps on all high, @ 1280x960! What a difference!
Im so happy now!
|
|
MaccerM
Member
|
21. August 2007 @ 04:35 |
Link to this message
|
|
Just out of interest Waymon which drivers were you using before that? I have catalyst 7.6 now but the ones off the CD that came with the card used to crash my machine quite alot, 7.6 was a big improvement.
Anyone else use those Omega drivers? I have read a few threads on them and it seems that the guy who makes those driver packs just mixes and matches files from different catalyst packs but doesn't test them on very many setups, consequently they're not too great. Apparantly Nvidia have asked him to stop doing it with their drivers because he makes such a hash of it!? I'll try and find the thread again...
Q9550 @ 3.4Ghz-1.275v & TRUE, Asus P5N-T Deluxe, 4Gb Dominator 1066 (5-5-5-15), Inno3d OC 8800GTs in SLi with TT Duorbs, 580w 5.1 Setup, 2x DVD-RAM, 320&160Gb SATA2 HDDs, 17" 3ms Hyundai TFT, 700w Jeantech PSU.
|
AfterDawn Addict
4 product reviews
|
21. August 2007 @ 04:44 |
Link to this message
|
|
Waymon: The case temperature is right, but it isn't the temperature of your case, no such sensor exists unless it's attached to some kind of external display, it's your GPU's ambient temperature, the temperature of everything on the board, to put it roughly.
950x2 = 1900, with the base frequency being 950. If all you can make is 855, that's 1.71Ghz memory frequency. I can only make 1.746, so you're not far off me.
The reason why the new Omega drivers work after that convoluted install process is probably because they don't recognise the AGP X1950XT as a supported ATi card (hardly surprising), but they do work with the card, and quite efficiently as you've pointed out. I think it's simply because the card is so unusual that the drivers aren't optimised for it, at all.
MaccerM: I used omega drivers for a while and had no issues, but didn't see that much benefit either. I think many people look to them because the default catalyst control center can be buggy.
|
|
MaccerM
Member
|
21. August 2007 @ 05:52 |
Link to this message
|
True CCC sux, but if you use it in conjunction with ATiTool then it's bareable. I just use CCC for all visual adjustments and display swapping, tool for the o/clocking and temp monitoring.
I can't be bothered with trying any of those omegas, they sound more trouble than they're worth! And feck it! I've been playing BF2 at 80-100 FPS in 1280x1024 in full with 6x AA, why the hell would I need to!?
Just need to get my cpu up to a good o/c now so I can play DiRT properly. It's in high gfx mode at 1024x768 and plays smooth with 30 fps on the rally tracks but starts to drop down to 15-20 on packed out races with lots of opponents. Don't ever think I've played such a drastically cpu limited game before! I've read that it's pretty much a straight 360 to pc port with no optimisation for the pc. Codemasters seem like they couldn't give a toss about pc users, not like the old days of CMR03 and 04.
Q9550 @ 3.4Ghz-1.275v & TRUE, Asus P5N-T Deluxe, 4Gb Dominator 1066 (5-5-5-15), Inno3d OC 8800GTs in SLi with TT Duorbs, 580w 5.1 Setup, 2x DVD-RAM, 320&160Gb SATA2 HDDs, 17" 3ms Hyundai TFT, 700w Jeantech PSU.
|
AfterDawn Addict
4 product reviews
|
21. August 2007 @ 06:01 |
Link to this message
|
|
That, I'm afraid is true of any game that started on the 360 first. Rainbow Six Vegas and Splinter Cell Double Agent are the same. They get put through an automatic code converter and they're not actually checked, just marketed, and if there are severe issues they release patches. Vegas is incredibly demanding, and I'm told that Double Agent is so full of bugs it's not worth the bother, fortunately we have the 360 version.
|
|
MaccerM
Member
|
21. August 2007 @ 07:01 |
Link to this message
|
Lets just hope GTA IV gets the proper treatment, Rockstar are pretty good at getting it right though. I'll just have to wait an extra 6 months for the PC version!
Q9550 @ 3.4Ghz-1.275v & TRUE, Asus P5N-T Deluxe, 4Gb Dominator 1066 (5-5-5-15), Inno3d OC 8800GTs in SLi with TT Duorbs, 580w 5.1 Setup, 2x DVD-RAM, 320&160Gb SATA2 HDDs, 17" 3ms Hyundai TFT, 700w Jeantech PSU.
|
Senior Member
|
21. August 2007 @ 07:41 |
Link to this message
|
Quote:
Just out of interest Waymon which drivers were you using before that?
I was using the drivers from the disc, which were the only ones I could find that didnt lag out with all of my games.
Quote:
but doesn't test them on very many setups,
I got this from the omegadrivers.net site:
Quote:
All Omega driver sets are tested (unless noted) by myself in my own PC or in an alternate PC (in the case of the Nvidia drivers) to ensure maximum compatibility and reliability.
Quote:
Waymon: The case temperature is right, but it isn't the temperature of your case, no such sensor exists unless it's attached to some kind of external display, it's your GPU's ambient temperature,
Sammorris, I've got a areocool coolwatch display in the front that has 3 temp probess hooked to it, 1 for the case, 1 for the HDD and 1 for the GFX card. When I said I get 86F tops in my case, that's what I was reading off of.
Quote:
o put it roughly.
950x2 = 1900,
So each time, Ijust double the base frequencey?
Quote:
Lets just hope GTA IV gets the proper treatment, Rockstar are pretty good at getting it right though. I'll just have to wait an extra 6 months for the PC version!
Yea, I want GTAIV too, but I would rather have my machine work ok with Crysis and Mercenaries 2. Sammorris says that my machine can do it, I just hope he's right!
BTW, I got this "physics emultion program" from here: http://www.cellfactorgame.com/forums/showthread.php?t=473
Does it work? I've been playing GRAW 2, the the hack that allows you to turn on the "extreme physics" without needing a PPU. I've just not really seen a difference...
|
AfterDawn Addict
4 product reviews
|
22. August 2007 @ 07:19 |
Link to this message
|
|
Hmm, look carefully for things that have extra physics, like being able to shoot holes in flags etc. If they're unchanged, the mod hasn't worked. Be advised, it places more of a strain on your graphics card as well.
|
Senior Member
|
22. August 2007 @ 10:53 |
Link to this message
|
Yup, scratch that last post. I can shoot a concrete wall, and the partacles come off and lay on the ground with the mod, they dont with out it. Also, when a grenade goes off, all these particles fly at you.
I can only play multiplayer with around 25-30fps, sometimes more. But when I play single player, forget it. The fps drop alot becasue of the bots I think.
|
AfterDawn Addict
4 product reviews
|
23. August 2007 @ 08:14 |
Link to this message
|
|
Yup, it is, and that's because you have a single core processor.
|
|
Advertisement
|
|
|
Senior Member
|
23. August 2007 @ 16:09 |
Link to this message
|
|
stupid old technology :(
Anyways, it is hyper threading, even though that doesnt do anything...
I wish I had a socket 775 mobo... Stupid old technology :-[
|
|
