1900XT vs GTX
- Thread starter alexsok
- Start date
geo said:
depends on how much stress the app is putting on it. Basically games where the video card is dishing out a good deal of effects (high textures, HDR, etc..) are going to push it into 60-75% RPM. It will keep relatively quiet with HDR off but full AA/AF in something like HL2 or its online offshoots, but with HDR on, or in something like the new Tomb Raider (full effects), its going to speed up to that 70% area. HDR tests in 3Dmark06 should give you an idea of the worst it will do. Assuming you can keep cool air blowing on the card you may not notice it at all which i dont. But not everyone has a big crossflow fan.
Spaceman-spiff: There's a review of the Blizzard X1900 XTX here if you want to take a look:
http://www.itreviews.co.uk/hardware/h937.htm
http://www.itreviews.co.uk/hardware/h937.htm
So I did some real-world tracking of this with Oblivion HDR/AA and looping the ice station HDR/AA scene in 3dm06.
My setup isn't the quietest in the first place because of my Akasa 650w psu. Not offensively loud, but noticeable. Airflow is very good with 3 120mm fans (one front, two back) and a Zalman CNPS9500 pushing air mid-case towards the back fans. And a side panel that has air inflow (Gigabyte 3DAurora).
What I found is below. The temp range is where the fan speed kicked in and where it went back down to the previous setting. This was tracked by playing on one 2405fpw while ATITool fanspeed tab and Overdrive temp monitor was up on the second monitor. x1900xt clocked at 641/792.
27% - 53c
31% - 70c - 67c
36% - 80c - 75c
50% - 88c - 85c
[Edit: Note this means there is overlap at a given temp point depending on whether temp is ramping or easing. 76c could be 31% if temps are on the way up. . .or 36% if they are on the way down from having been 80c or above]
I never had it go above 50% on fanspeed as measured by ATITool. 50% never failed to immediately begin dropping the temp. So in hard usage (in my environment anyway), the variability of noticeable noise was that cycling between 36% and 50% fanspeed. As I think I said above, pretty much anything above 45% is noticeable and annoying because of the whine associated. But I didn't notice it when playing Oblivion, presumably because of the game sound and the focus on playing. Did notice it more with 3dm06 because neither is present.
So. Still haven't ordered a 3rd party cooler. Still haven't decided not to either. We'll see what more experience brings. I may try locking it at 45%. Originally I assumed that would be much too low for heavy gaming, but since I haven't seen it go above 50% on its own yet, and haven't seen 50% fail to lower temps, it could be that a constant 45% would be acceptable.
My setup isn't the quietest in the first place because of my Akasa 650w psu. Not offensively loud, but noticeable. Airflow is very good with 3 120mm fans (one front, two back) and a Zalman CNPS9500 pushing air mid-case towards the back fans. And a side panel that has air inflow (Gigabyte 3DAurora).
What I found is below. The temp range is where the fan speed kicked in and where it went back down to the previous setting. This was tracked by playing on one 2405fpw while ATITool fanspeed tab and Overdrive temp monitor was up on the second monitor. x1900xt clocked at 641/792.
27% - 53c
31% - 70c - 67c
36% - 80c - 75c
50% - 88c - 85c
[Edit: Note this means there is overlap at a given temp point depending on whether temp is ramping or easing. 76c could be 31% if temps are on the way up. . .or 36% if they are on the way down from having been 80c or above]
I never had it go above 50% on fanspeed as measured by ATITool. 50% never failed to immediately begin dropping the temp. So in hard usage (in my environment anyway), the variability of noticeable noise was that cycling between 36% and 50% fanspeed. As I think I said above, pretty much anything above 45% is noticeable and annoying because of the whine associated. But I didn't notice it when playing Oblivion, presumably because of the game sound and the focus on playing. Did notice it more with 3dm06 because neither is present.
So. Still haven't ordered a 3rd party cooler. Still haven't decided not to either. We'll see what more experience brings. I may try locking it at 45%. Originally I assumed that would be much too low for heavy gaming, but since I haven't seen it go above 50% on its own yet, and haven't seen 50% fail to lower temps, it could be that a constant 45% would be acceptable.
I hear the XTX I have installed spinning up and down while playing Battle for Middle-earth 2. But either I'm cheap or I'm not too annoyed by it, because I haven't gone the 3rd-party cooling solution either.
The shame of it is, my sense now is that they really aren't all that far off from having a reasonable solution. 50% (or maybe 55% for one more bump up for people with a little less air flow in their cases) of the current solution ought to be set to be 100%. And then a bit better quality fan motor that won't whine for that last jump to the new 100% (old 50-55%).
I do think part of the perception problem is OCer and enthusiast types getting into ATITool and setting that fan up above where it would be in real world for a stock configuration and having it sound really, really bad (and it does to my ear). Which, mea culpa, was the first thing I did too. But if the current 50% sounded like the current 36% (possibly a touch more "moving air" noise, but not that motor whine), and the current 50% was as high as you could set it (again, making it effectively 100% in the new regime), I doubt you'd hear any bitching about the noise on this thing.
I do think part of the perception problem is OCer and enthusiast types getting into ATITool and setting that fan up above where it would be in real world for a stock configuration and having it sound really, really bad (and it does to my ear). Which, mea culpa, was the first thing I did too. But if the current 50% sounded like the current 36% (possibly a touch more "moving air" noise, but not that motor whine), and the current 50% was as high as you could set it (again, making it effectively 100% in the new regime), I doubt you'd hear any bitching about the noise on this thing.
geo said:
Alot of cards have fans that go to pretty ridiculous levels at 100% though Geo. Why should this one be the exception with a restricted level? Did you ever try to push an X800 cooler to 100%? It sounds like its rattling to peices. The same goes for the coolers on 7800s. Mine does kick into the high 70% range to keep it at 80C if i shut off all my case cooling, so ATI didnt really do anything wrong in my opinion with how it linked RPM to thermal output.
Jawed
Legend
Uninstalls Oblivion, Tomb Raider: Legends, FEAR, etc.
Phew, thx Jawed.
Phew, thx Jawed.
Bouncing Zabaglione Bros.
Legend
FiringSquad reviewed Zalman VF900-Cu on X1900XTX...
http://www.firingsquad.com/hardware/zalman_vf900-cu/
http://www.firingsquad.com/hardware/zalman_vf900-cu/
FiringSquad said:
Bouncing Zabaglione Bros. said:
Oh, I do like the looks of that. :smile: That was one of my first thoughts, to wonder if putting the fan at the back instead of right next to the grill to the outside wouldn't help a bit too. Put more. . .stuff. . .between the noisemaker and the outside.
Now if we can just get them to sell the cooler aftermarket. . .
And better yet, get ATI to adopt it for a reference cooler.Bouncing Zabaglione Bros. said:
The problem is, the HIS X1900 you can actually buy at Newegg right now has the reference cooler.
This is part of the problem (I wanted to say "disingenuousness", but decided to be generous on a Friday afternoon) with Bergman's response that it was up to AIB's to put better coolers on. We all know that the first high-end cards on the market are going to be single-source reference models with possibly an AIB sticker on it --and that's it. So your hardest hardcore loyalists who are begging to put their disposable income in your pocket are left with the choice of getting a cooler they won't like, or waiting for later revs. That's not good for anybody, in my book.
Similar threads
