Tonedef
10-30-2007, 07:18 PM
Since the other post seems broken :)
http://webpages.charter.net/bliss/
http://webpages.charter.net/bliss/
View Full Version : Crysis Demo: DirectX 9 vs. DirectX 10 screenshots
|
|
j00st81
10-30-2007, 11:58 PM
http://www.imagebam.com/image/3a8c15609402
mabye its good that my comp can't run it like that, i'd be staring at the enviroments too much to notice all the enemies shooting at me :o
mabye its good that my comp can't run it like that, i'd be staring at the enviroments too much to notice all the enemies shooting at me :o
doCHtor
10-31-2007, 07:46 AM
It's possible to run Crysis in "DX10 graphics" (very high details) on dx9/winxp too. All you need to do is go to "Crysis SP Demo\Game\Config\CVarGroups" and in every cfg file copy the settings from default (very high detail) to the section which is marked as [3] (high detail) and now if you setup crysis (in ingame menu) to high detail, it will actually run on very high details..... so much for vista and dx10 :scream:
richcz3
10-31-2007, 08:21 AM
The DX10 expections are pretty over rated. Crytek played it very smart by not alienating a huge audience for such a high end PC only game (Won't see a console port).
On a few sites where there where side by side comparisons, the DX10 difference is very subtle. The hardware perfomance hit of going to Vista for DX10 is hard to qualify.
The only thing I read that Cervat Yerli said that struck me as odd, had to do with multiplayer. People with DX9 versions of the game will not be able to join servers running the DX10 versions of the game. That there are some gameplay dynamics in DX10 that prohibit DX9 players from joining. :shrug:
On a few sites where there where side by side comparisons, the DX10 difference is very subtle. The hardware perfomance hit of going to Vista for DX10 is hard to qualify.
The only thing I read that Cervat Yerli said that struck me as odd, had to do with multiplayer. People with DX9 versions of the game will not be able to join servers running the DX10 versions of the game. That there are some gameplay dynamics in DX10 that prohibit DX9 players from joining. :shrug:
ringzero
10-31-2007, 11:05 AM
the DX10 difference is very subtle. The hardware perfomance hit of going to Vista for DX10 is hard to qualify. :shrug:
From the images I`d call the differences quite large... your a hard man to please 8)
From the images I`d call the differences quite large... your a hard man to please 8)
BoostAbuse
10-31-2007, 05:12 PM
The only thing I read that Cervat Yerli said that struck me as odd, had to do with multiplayer. People with DX9 versions of the game will not be able to join servers running the DX10 versions of the game. That there are some gameplay dynamics in DX10 that prohibit DX9 players from joining. :shrug:
It's mainly due to competitive reasons, the performance difference between the DX10 MP and DX9 MP is ridiculous... DX10 on full specs with an 8800GTX still cripples the card and SLI isn't fully optimized yet even with Crysis being the big "push" for Nvidia and SLI. To make an example, if I run 3dmark06 on my machine I score almost 16,000 points in total... if I run 3dmark05 which is DX9 I practically double that.
Another factor is that they use the DX10 engine for doing the destructible environments through the GPU for effects such as blowing up trees etc (GPU killer!) so they've also said it's a gameplay thing of sorts as well... begs the question though about what happens when they release the DX10.1 patch, now we get 3 groups of servers?
It's mainly due to competitive reasons, the performance difference between the DX10 MP and DX9 MP is ridiculous... DX10 on full specs with an 8800GTX still cripples the card and SLI isn't fully optimized yet even with Crysis being the big "push" for Nvidia and SLI. To make an example, if I run 3dmark06 on my machine I score almost 16,000 points in total... if I run 3dmark05 which is DX9 I practically double that.
Another factor is that they use the DX10 engine for doing the destructible environments through the GPU for effects such as blowing up trees etc (GPU killer!) so they've also said it's a gameplay thing of sorts as well... begs the question though about what happens when they release the DX10.1 patch, now we get 3 groups of servers?
richcz3
10-31-2007, 06:05 PM
From the images I`d call the differences quite large... your a hard man to please 8)ringzero - I should clear up my point a bit. doCHtor mentions the mod edit way to increase the effects for use for DX9 cards. By the time the full game is out, I'm sure there will be new methods to max effects on DX9 hardware. Forcing DX9 to do the effects in XP vs. using dedicated DX10 Hardware in Vista. There's an interesting trade off there.
With all that said - the Crysis engine takes some serious hardware regardless if you're DX9 or DX10. DX10 hardware in Vista requires even more resources. Increased effects on a DX9 in XP will punish anything but an uber spec'd system.
BoostAbuse - I didn't play the MP beta, but everything I've read, it requires a high end system to get decent visuals/frame rates in DX9. What exactly will with DX10 and Vista require? Cervat said they shot for wide scaleability - trick is if gamers are willing to water down the effects just to play. Interesting you mention benchmarks as I am betting Crysis will be an official DX10 staple for hardware sites.
With all that said - the Crysis engine takes some serious hardware regardless if you're DX9 or DX10. DX10 hardware in Vista requires even more resources. Increased effects on a DX9 in XP will punish anything but an uber spec'd system.
BoostAbuse - I didn't play the MP beta, but everything I've read, it requires a high end system to get decent visuals/frame rates in DX9. What exactly will with DX10 and Vista require? Cervat said they shot for wide scaleability - trick is if gamers are willing to water down the effects just to play. Interesting you mention benchmarks as I am betting Crysis will be an official DX10 staple for hardware sites.
doCHtor
10-31-2007, 06:23 PM
I will just add that so far it seems that there is no, or almost no visual difference between dx9 crysis which runs on very high detail (through cfg tweaks) and dx10 crysis which runs on very high details without any tweaks.
Interesting is, that it also seems that dx9 with very high details doesn't run slower than very high det. on dx10. Or let's say drop in fps on dx9 (high vs. very high detail) is not bigger than drop in fps on dx10.
Interesting is, that it also seems that dx9 with very high details doesn't run slower than very high det. on dx10. Or let's say drop in fps on dx9 (high vs. very high detail) is not bigger than drop in fps on dx10.
ringzero
10-31-2007, 11:27 PM
I get you - It would be interesting to see DX9 cfg tweaked comparisson images.
doCHtor
11-01-2007, 10:09 AM
Here you go:
http://www.incrysis.com/forums/viewtopic.php?id=11451
http://www.incrysis.com/forums/viewtopic.php?id=11451
daHund
11-01-2007, 01:37 PM
[...]
The only thing I read that Cervat Yerli said that struck me as odd, had to do with multiplayer. People with DX9 versions of the game will not be able to join servers running the DX10 versions of the game. That there are some gameplay dynamics in DX10 that prohibit DX9 players from joining. :shrug:
thats actually one of cryteks remarkable marketing tricks ;) - to push technology of their investors. they did the same with their 64bit version of farcry.
the reason dx9 gamers can't play on dx10 servers is simply physics. while all crysis physics are cpu based, dx10 players are suspected to have better cpus installed - so dx10 servers run much better physics simulations (destructible environments, mainly).
this of course is complete nonsense, because even if users with dx9 hardware mostly didn't have up to date cpus, dx10 is only available to vista users. so there is a fair amount of dx10 hardware running dx9.
it is for reasons like that, why you shouldn't buy german computer games ;) (well, shooters. there are some very nice german adventure games, go buy those instead :P). those decisions are clearly not in favor of gamers - so, as a gamer, don't support them!
P.S.: this post is written by a german citizen, so don't take it as racism ;).
P.P.S.: for some insight into the "very high" marketing gag of crysis, look here (http://www.digitalreport.net/content/view/64/1/1/0/).
The only thing I read that Cervat Yerli said that struck me as odd, had to do with multiplayer. People with DX9 versions of the game will not be able to join servers running the DX10 versions of the game. That there are some gameplay dynamics in DX10 that prohibit DX9 players from joining. :shrug:
thats actually one of cryteks remarkable marketing tricks ;) - to push technology of their investors. they did the same with their 64bit version of farcry.
the reason dx9 gamers can't play on dx10 servers is simply physics. while all crysis physics are cpu based, dx10 players are suspected to have better cpus installed - so dx10 servers run much better physics simulations (destructible environments, mainly).
this of course is complete nonsense, because even if users with dx9 hardware mostly didn't have up to date cpus, dx10 is only available to vista users. so there is a fair amount of dx10 hardware running dx9.
it is for reasons like that, why you shouldn't buy german computer games ;) (well, shooters. there are some very nice german adventure games, go buy those instead :P). those decisions are clearly not in favor of gamers - so, as a gamer, don't support them!
P.S.: this post is written by a german citizen, so don't take it as racism ;).
P.P.S.: for some insight into the "very high" marketing gag of crysis, look here (http://www.digitalreport.net/content/view/64/1/1/0/).
Rickmeister
11-01-2007, 02:35 PM
This game is not playable on high resolutions! on a 1920x1200 resolution there is no chance I can run on a 'very high detail', even high is causing massive fps lag. So that sucks pretty bigtime.
btw, i'm running on a 8800gtx 640mb...
btw, i'm running on a 8800gtx 640mb...
Tamis
11-01-2007, 03:36 PM
you got dx9 10 specific servers ???
DX is purly a visual driver, dynamics have nothing to do with DX.
i gues microsoft has a hand in this tryng to force people to switch and buy vista.
DX is purly a visual driver, dynamics have nothing to do with DX.
i gues microsoft has a hand in this tryng to force people to switch and buy vista.
Koogle
11-01-2007, 03:50 PM
I suppose the dx10 one is perhaps mslighty more realistic but I actually prefere the dx9 colours, ... although the dof blur in the dx9 isn't as good as the dx10
BoostAbuse
11-01-2007, 04:11 PM
you got dx9 10 specific servers ???
DX is purly a visual driver, dynamics have nothing to do with DX.
i gues microsoft has a hand in this tryng to force people to switch and buy vista.
incorrect actually, if you take a look at how they are doing their environmental effects it's mostly GPU driven. DX 10 opens up a lot of new development both visually and on other areas allowing for some very interesting results. I.e. cloth simulation on the gpu, skinning on the gpu (both spherical and linear) as well as deploying some additional algorithms for doing destruction of the trees :)
Rich: DX10 and Vista require a bit heavier of a system spec than the DX9 counterpart. The DX9 version runs smooth on my machine but I run a maximum resolution of 1440x900 which is a sweet spot for the Radeon's. As others have said, running this game at full quality settings in both DX9 and DX10 at 1920x1200 is just a nightmare even for a 8800GTX card and without proper SLI support until Nvidia fixes up their drivers some more people will certainly water down their settings in order to be competitive online. Remember as well that the 8800GTS 640mb cards are great up until about 1600x1200 at which point they begin to lose their luster.
As for the DX10 part, I can run it on my machine and get decent visuals in SP but with the MP aspect I doubt it will perform up to par when you factor in all the other players and dynamics of online play. John Carmack was commenting on the whole Crysis thing and he said at bare minimum users should have a GeForce 8 series with 512mb of RAM minimum if they wish to play at 1600x1200... he also hinted towards the 8800GT series which seems to handle Crysis on par with an 8800GTX on high settings (single player).
Try running a Crysis techdemo on your machine and see how your results fare compared to others with the various DX flavours.
DX is purly a visual driver, dynamics have nothing to do with DX.
i gues microsoft has a hand in this tryng to force people to switch and buy vista.
incorrect actually, if you take a look at how they are doing their environmental effects it's mostly GPU driven. DX 10 opens up a lot of new development both visually and on other areas allowing for some very interesting results. I.e. cloth simulation on the gpu, skinning on the gpu (both spherical and linear) as well as deploying some additional algorithms for doing destruction of the trees :)
Rich: DX10 and Vista require a bit heavier of a system spec than the DX9 counterpart. The DX9 version runs smooth on my machine but I run a maximum resolution of 1440x900 which is a sweet spot for the Radeon's. As others have said, running this game at full quality settings in both DX9 and DX10 at 1920x1200 is just a nightmare even for a 8800GTX card and without proper SLI support until Nvidia fixes up their drivers some more people will certainly water down their settings in order to be competitive online. Remember as well that the 8800GTS 640mb cards are great up until about 1600x1200 at which point they begin to lose their luster.
As for the DX10 part, I can run it on my machine and get decent visuals in SP but with the MP aspect I doubt it will perform up to par when you factor in all the other players and dynamics of online play. John Carmack was commenting on the whole Crysis thing and he said at bare minimum users should have a GeForce 8 series with 512mb of RAM minimum if they wish to play at 1600x1200... he also hinted towards the 8800GT series which seems to handle Crysis on par with an 8800GTX on high settings (single player).
Try running a Crysis techdemo on your machine and see how your results fare compared to others with the various DX flavours.
bluecanvas
11-01-2007, 05:10 PM
incorrect actually, if you take a look at how they are doing their environmental effects it's mostly GPU driven. DX 10 opens up a lot of new development both visually and on other areas allowing for some very interesting results. I.e. cloth simulation on the gpu, skinning on the gpu (both spherical and linear) as well as deploying some additional algorithms for doing destruction of the trees :)
I believe that Nvidia/ATI may be omitting functionality from their XP drivers to make it look as if certain GPU functions can only be accessed under the Vista/DX10 driver model.
If this is the case and someone proves it, nobody will believe a word ATI/Nvidia/Microsoft have to say about gaming graphics again.
I believe that Nvidia/ATI may be omitting functionality from their XP drivers to make it look as if certain GPU functions can only be accessed under the Vista/DX10 driver model.
If this is the case and someone proves it, nobody will believe a word ATI/Nvidia/Microsoft have to say about gaming graphics again.
kees
11-01-2007, 05:32 PM
incorrect actually, if you take a look at how they are doing their environmental effects it's mostly GPU driven. DX 10 opens up a lot of new development both visually and on other areas allowing for some very interesting results. I.e. cloth simulation on the gpu, skinning on the gpu (both spherical and linear) as well as deploying some additional algorithms for doing destruction of the trees :)
Those are still only visual things and do not affect (afaik) any simulation happening in the game. Unless you are saying they are sending the cloth simulated mesh back to the engine.
Which would be the first time I've heard of it (not saying its not possible, Crytek knows better then me no doubt).
Usually whatever happens on the GPU, stays on the GPU (kinda like Las Vegas ;P ) and is never pulled from the GPU back into the CPU to affect other simulations that could affect gameplay or multiplayer etc.
Those are still only visual things and do not affect (afaik) any simulation happening in the game. Unless you are saying they are sending the cloth simulated mesh back to the engine.
Which would be the first time I've heard of it (not saying its not possible, Crytek knows better then me no doubt).
Usually whatever happens on the GPU, stays on the GPU (kinda like Las Vegas ;P ) and is never pulled from the GPU back into the CPU to affect other simulations that could affect gameplay or multiplayer etc.
richcz3
11-01-2007, 06:40 PM
i......Try running a Crysis techdemo on your machine and see how your results fare compared to others with the various DX flavours.
I'm prepping a quadcore (3GHz) today and hope to get a GT or GTS in it before tomorrow. I'll compare this to the duals (3.4GHz) running GTS and ATIx1950s. I don't use Vista so the best I can do is compare in XP.
I'm thinking the processors play a bigger role than Crytek may be leading on. If PC gaming has a leg to stand on its multicore processing. I'm sure that DX10 (GPU) plays a large role but Vista + DX10 Uber PCs must be an infinitely small market share compared to DX9/XP. That and most FPS multiplayer games reduce dynamics available in SP to appeal to a wider hardware audience.
I know that Far Cry had a solid multiplayer game but its system specs (at the time) reduced online games to a trickle. Its unfortunate because Coop levels were possible.
Having heard the MP Beta was demanding, I can't help but feel that Crysis won't be relegated to the hardware site benchmark statistics after people finish the single player.
I'm prepping a quadcore (3GHz) today and hope to get a GT or GTS in it before tomorrow. I'll compare this to the duals (3.4GHz) running GTS and ATIx1950s. I don't use Vista so the best I can do is compare in XP.
I'm thinking the processors play a bigger role than Crytek may be leading on. If PC gaming has a leg to stand on its multicore processing. I'm sure that DX10 (GPU) plays a large role but Vista + DX10 Uber PCs must be an infinitely small market share compared to DX9/XP. That and most FPS multiplayer games reduce dynamics available in SP to appeal to a wider hardware audience.
I know that Far Cry had a solid multiplayer game but its system specs (at the time) reduced online games to a trickle. Its unfortunate because Coop levels were possible.
Having heard the MP Beta was demanding, I can't help but feel that Crysis won't be relegated to the hardware site benchmark statistics after people finish the single player.
BoostAbuse
11-01-2007, 07:47 PM
A quick suggestion if you are setting up a Q6600 and something that I've noticed that you clued into as well is that the FSB plays a very large factor in all of this. One modification I made to my rig was switching the Q6600 from 266x9 to 333x9 but in doing so I noticed that my CPU:Memory ratio was slightly off causing some really interesting latency issues. I re-adjusted it this week to run 400x8 (3.2ghz) which gave me 1:1 on DDR2-6400 and my memtest results had next to no latency issues running factory voltage and factory 5-5-5-12 timings. The 400x8 bumped my FSB to 1600mhz up from the stock 1066 and the performance improvement this made was quite astounding.
richcz3
11-01-2007, 09:03 PM
A quick suggestion if you are setting up a Q6600 and something that I've noticed that you clued into as well is that the FSB plays a very large factor in all of this.....Spot on - the Q6600 it is. Paired with EVGA/6801 motherboard.
Thanks for the settings - Right now running conservative - x9 1451FSB Unlinked with Mem at 800 and get 3.2Ghz. Crucial Ballistix is stock speed @ 800. Just started the Orthos test and will let it run for a day to check for errors. This Q6600 is the low TDP part which is noted for some high OC. So theres room to bump it up if this proves cool and stable.
EDIT - Forget the bump - Throwing caution to the wind Running Orthos x2 @ 3.5GHz.
Amazingly cool with four cores @ 100% 52-58 load temps
Thanks for the settings - Right now running conservative - x9 1451FSB Unlinked with Mem at 800 and get 3.2Ghz. Crucial Ballistix is stock speed @ 800. Just started the Orthos test and will let it run for a day to check for errors. This Q6600 is the low TDP part which is noted for some high OC. So theres room to bump it up if this proves cool and stable.
EDIT - Forget the bump - Throwing caution to the wind Running Orthos x2 @ 3.5GHz.
Amazingly cool with four cores @ 100% 52-58 load temps
BoostAbuse
11-02-2007, 02:00 PM
Spot on - the Q6600 it is. Paired with EVGA/6801 motherboard.
Thanks for the settings - Right now running conservative - x9 1451FSB Unlinked with Mem at 800 and get 3.2Ghz. Crucial Ballistix is stock speed @ 800. Just started the Orthos test and will let it run for a day to check for errors. This Q6600 is the low TDP part which is noted for some high OC. So theres room to bump it up if this proves cool and stable.
EDIT - Forget the bump - Throwing caution to the wind Running Orthos x2 @ 3.5GHz.
Amazingly cool with four cores @ 100% 52-58 load temps
Not bad results, this weekend I will be attempting 400x9 to make 3.6ghz on air and maintain the 1:1 memory ratio (that's if my transistor caps don't blow off the motherboard first). What cooling are you using on the machine? By the sounds of it you got yourself a nice G0 model, not the crap B3 model :) Are those temps monitored via CoreTemp or SpeedFan? SpeedFan's got that crappy issue about underreporting temperatures by 15 degrees...
Thanks for the settings - Right now running conservative - x9 1451FSB Unlinked with Mem at 800 and get 3.2Ghz. Crucial Ballistix is stock speed @ 800. Just started the Orthos test and will let it run for a day to check for errors. This Q6600 is the low TDP part which is noted for some high OC. So theres room to bump it up if this proves cool and stable.
EDIT - Forget the bump - Throwing caution to the wind Running Orthos x2 @ 3.5GHz.
Amazingly cool with four cores @ 100% 52-58 load temps
Not bad results, this weekend I will be attempting 400x9 to make 3.6ghz on air and maintain the 1:1 memory ratio (that's if my transistor caps don't blow off the motherboard first). What cooling are you using on the machine? By the sounds of it you got yourself a nice G0 model, not the crap B3 model :) Are those temps monitored via CoreTemp or SpeedFan? SpeedFan's got that crappy issue about underreporting temperatures by 15 degrees...
richcz3
11-02-2007, 10:15 PM
8 hours and boom. No dice. 3.5Ghz didnt take for me.
Its not a thermal issue either. I add voltage without a significant heat penalty, but it doens't stick. So I throttled down to 3.2Ghz and shes holding now.
So I did a 2core vs 4 core comparison
I have two systems identical xppro/3.2GHz/2gig-800mem/ATIx1950Pro
@ 1280x1024.
I can't see a discernable difference between the two systems. I don't know how to get bring up the FPS meter in game to detect the quads benefit over the dual core.
Its not a thermal issue either. I add voltage without a significant heat penalty, but it doens't stick. So I throttled down to 3.2Ghz and shes holding now.
So I did a 2core vs 4 core comparison
I have two systems identical xppro/3.2GHz/2gig-800mem/ATIx1950Pro
@ 1280x1024.
I can't see a discernable difference between the two systems. I don't know how to get bring up the FPS meter in game to detect the quads benefit over the dual core.
BoostAbuse
11-02-2007, 11:45 PM
8 hours and boom. No dice. 3.5Ghz didnt take for me.
Its not a thermal issue either. I add voltage without a significant heat penalty, but it doens't stick. So I throttled down to 3.2Ghz and shes holding now.
So I did a 2core vs 4 core comparison
I have two systems identical xppro/3.2GHz/2gig-800mem/ATIx1950Pro
@ 1280x1024.
I can't see a discernable difference between the two systems. I don't know how to get bring up the FPS meter in game to detect the quads benefit over the dual core.
You probably won't notice much of a difference unless the game or application is threaded to support more than 2 processors or 4 cores.
As far as making it stick at 3.5ghz, check your Northbridge voltage and maybe raise it a bit. The NB is the weak spot at higher clock speeds and is generally the first thing to give way depending on the board you use. I'll be trying for 3.6 this weekend though I highly doubt my P5K will handle that :)
Its not a thermal issue either. I add voltage without a significant heat penalty, but it doens't stick. So I throttled down to 3.2Ghz and shes holding now.
So I did a 2core vs 4 core comparison
I have two systems identical xppro/3.2GHz/2gig-800mem/ATIx1950Pro
@ 1280x1024.
I can't see a discernable difference between the two systems. I don't know how to get bring up the FPS meter in game to detect the quads benefit over the dual core.
You probably won't notice much of a difference unless the game or application is threaded to support more than 2 processors or 4 cores.
As far as making it stick at 3.5ghz, check your Northbridge voltage and maybe raise it a bit. The NB is the weak spot at higher clock speeds and is generally the first thing to give way depending on the board you use. I'll be trying for 3.6 this weekend though I highly doubt my P5K will handle that :)
PhantomDesign
11-03-2007, 09:14 AM
I have a q6600 in my computer @ 3.3ghz, 100% stable, runs VERY cool. I suppose it helps that my case is loaded with 3x 120mm fans, 1x 200mm fan, Thermalite 120, lapped heatsink & processor (sanded them FLAT), etc.
I could push it a lot more than 3.3, but I don't feel like pushing the voltages higher & I'm happy with my .9 ghz boost for now.
---
Anyone else try out the editor? It's really easy to use; wish I could export directly to 3DS max.
Here's a manual that covers basic functions . .
http://doc.crymod.com/frames.html?frmname=topic&frmfile=Voxel.html
I could push it a lot more than 3.3, but I don't feel like pushing the voltages higher & I'm happy with my .9 ghz boost for now.
---
Anyone else try out the editor? It's really easy to use; wish I could export directly to 3DS max.
Here's a manual that covers basic functions . .
http://doc.crymod.com/frames.html?frmname=topic&frmfile=Voxel.html
BoostAbuse
11-03-2007, 02:11 PM
3.2-3.3 is a nice stable point on the Q6600 :) I finally hit 3.6ghz on air today, what a battle that was. http://service.futuremark.com/compare?3dm06=3621491 That's at 1.48v (I can hear my motherboard screaming for mercy) 400x9.. Today will be the day I think I'll change the system over to a water/tec based cooler to keep temps under control as during CPU Test of 3dmark06 I was pushing 62-67 degrees on the CPU which is not cool
----
took a rummage through the editor docs and also have all of crytek's presentation stuff from GDC 2007, I can't argue that these guys are damn smart with what they've done with the engine but man... you really kill your bottom line targeting just 'uber' gamers and limiting yourself from consoles..
----
took a rummage through the editor docs and also have all of crytek's presentation stuff from GDC 2007, I can't argue that these guys are damn smart with what they've done with the engine but man... you really kill your bottom line targeting just 'uber' gamers and limiting yourself from consoles..
richcz3
11-05-2007, 01:50 AM
Persistance paid off Hit 3.5Ghz on Friday night. Stress tested Saturday.
Using EVGA 680i motherboard which is real particular but dialed in the perfect fsb Linked/Sync @ 390x9. Temps are just tolerable under load so I'm sticking with it. All High settings default using the 8800GTS - much smoother now
BoostAbuse - 3.2 - 3.3 are the sweet spot for this chip. Anything above is hit or miss as I've been reading on some forums.
PhantomDesign - What board are you using?
Using EVGA 680i motherboard which is real particular but dialed in the perfect fsb Linked/Sync @ 390x9. Temps are just tolerable under load so I'm sticking with it. All High settings default using the 8800GTS - much smoother now
BoostAbuse - 3.2 - 3.3 are the sweet spot for this chip. Anything above is hit or miss as I've been reading on some forums.
PhantomDesign - What board are you using?
PhantomDesign
11-05-2007, 01:56 AM
[FULL COMPUTER BUILD LINK] (http://secure.newegg.com/NewVersion/wishlist/PublicWishDetail.asp?WishListNumber=8040507)
I got that board specifically for the P35 chipset, which is supposed to be great for quad-cores.
I got that board specifically for the P35 chipset, which is supposed to be great for quad-cores.
CGTalk Moderation
11-05-2007, 01:56 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.
vBulletin v3.0.5, Copyright ©2000-2012, Jelsoft Enterprises Ltd.