980 Ti

OwNjaH said:
B&H has the normal 980ti ACX2.0+ for 649 plus free shipping right now. Thinking of pulling the trigger on that. The 980 FTW I have now didn't really live up to my expectations (ASIC at 65%), so I think I just want to get the base model and OC that.

Edit: Has anyone purchased something from B&H? How is their refund policy? After my latest experience with Newegg I'm open to buying stuff from other places.
Click to expand...
They're great. Had to swap out a defective camera and they gave me an advance exchange.

Probably a restocking fee if it's not defective though. 30-day return window is pretty standard.

Interestingly TechPowerUp found the 980 Ti Gaming to be noisier than the ACX 2.0:
http://www.techpowerup.com/mobile/reviews/MSI/GTX_980_Ti_Gaming/29.html

Based on my experience I wouldn't be surprised if the ACX 2.0 was actually the quietest of the bunch (contrary to expectations).
 
MSI boosts a bit higher and runs about 2C cooler.

Both EVGA and MSI are re-using the exact same heatsinks they used on their 980 models. If the MSI 980 TF5 is superior to the 980 ACX 2.0 (not +) then the Ti version should be the same way.

ACX has always been far-and-away the WORST heatsink design of all major Nvidia partners.
In this case, it looks like the EVGA model uses a more robust heatsink: More heatpipes, horizontal layout.

cooler2.jpg


cooler2.jpg


card3.jpg


card3.jpg
 
Last edited:
friend'scatdied said:
They're great. Had to swap out a defective camera and they gave me an advance exchange.

Probably a restocking fee if it's not defective though. 30-day return window is pretty standard.
Click to expand...

I spoke with a rep when I thought about buying from them, they have no restocking fee within the 30 day window. If you don't like it, send it back and they will even provide shipping label, no fee. Newegg could learn from them
 
TaintedSquirrel said:
MSI boosts a bit higher and runs about 2C cooler.
Click to expand...
Actually it's boosting a lot higher out-of-box: [email protected] (vs. [email protected] on the eVGA). Interestingly the load difference at max overclocks is still 2'C (MSI@1507 vs. eVGA@1491).

I still think the noise performance is a surprising upset since the eVGA ACX2.0+ is basically the same size as the reference cooler (vs. other designs being 30% larger). The fact that MSI/ASUS/et al. are throwing such huge coolers on their cards for a 2'C difference (with louder fan speeds) seems to be a disappointment.
 
Last edited:
Ah, not in the main description, gotcha
Thats a good deal... If only I didn't want something cooler....
 
Dissapointed in this cards. I didn't know i needed an X99 setup to get good high frames.

Everything is choppy as hell, 4770k is huge bottleneck for the setup. If I want to game i have to run MSAA 4x, otherwise it's all choppy with 55-60% gpu usage per card, embarassing.

Ooh, and the firestrike test 4, combined, is god awfull. Only x99 scores good there. Sucks

1k corsair psu also not coping (it has degraded over all these years and it's 2x500W psu inside actually i think, new 1000w should not be any problem), i tried 1450 oc and nearly blew out, insta shutoff (works good up untill 1380). And they are showing 225-260W consumption with some even recommending 700W psu for sli of these :)
 
Last edited:
AriciU said:
Dissapointed in this cards. I didn't know i needed an X99 setup to get good high frames.

Everything is choppy as hell, 4770k is huge bottleneck for the setup. If I want to game i have to run MSAA 4x, otherwise it's all choppy with 55-60% gpu usage per card, embarassing.

Ooh, and the firestrike test 4, combined, is god awfull. Only x99 scores good there. Sucks

1k corsair psu also not coping (it has degraded over all these years and it's 2x500W psu inside actually i think, new 1000w should not be any problem), i tried 1450 oc and nearly blew out, insta shutoff (works good up untill 1380). And they are showing 225-260W consumption with some even recommending 700W psu for sli of these :)
Click to expand...

Everything in this post is wrong and either issues with a specific game or poster's system. A 4770k is not going to bottleneck the card at all.

A 980 Ti requires a minimum 600W PSU, but probably 750W is ideal. Of course it needs to have enough rails and whatnot to provide stable power to the card. For SLI you want 1000W PSU at least. Try again when you get a properly working PSU as getting computer to shut off when overclocked does not sound good.
 
I will believe that when i see another 4770k with it working properly.

So it is normal running like crap with AA off while enabling it magically solves the problem?
 
AriciU said:
Care to point one out so that i may ask him? Its driving me insane. Z87 board.
Click to expand...

im running it with a 3770K.. (~5-10% above 2600K so kinda comparable 99% of the time) no single issues, no bottlenecks.. so you probably have another kind of specific game issues, in the other hand have you updated your BIOS? have you tested your cards individually?.
 
Araxie said:
im running it with a 3770K.. (~5-10% above 2600K so kinda comparable 99% of the time) no single issues, no bottlenecks.. so you probably have another kind of specific game issues, in the other hand have you updated your BIOS? have you tested your cards individually?.
Click to expand...

Great, i see that you are running a Z77 board. You are running x8 x8 also right? What are you results in firestrike test 4? Can you look at you gpu usage % from afterburner osd during that specific test please?

I have tried both cards individually extensively trying to find max oc clocks for each. There were no problems with either. One can do 1493 boost, the other about 1480 (3d mark bench stable, not game stable). I have the latest bios on my z87x-d3h gigabyte board.

I will repost some screenshots showing my issues, maybe that will explain it better.

surfhick said:
misterbobby, in this thread

http://hardforum.com/showthread.php?t=1864094&page=24

is running an overclocked 4770. His results are actually making me feel good about my purchase.
Click to expand...

Thanks for that. I will investigate further. Is he running SLI though? I see only one card in his signature.
 
AriciU said:
I have tried both cards individually extensively trying to find max oc clocks for each. There were no problems with either. One can do 1493 boost, the other about 1480 (3d mark bench stable, not game stable). I have the latest bios on my z87x-d3h gigabyte board.
Click to expand...

Are you causing your card(s) to throttle by maxing out the OC?
 
I took some screenshots. They are 1440p res so will put links to not mess everything up with the IMG function.

Firestrike test1, outlined is the GPU usage % - http://i.imgur.com/myICBGu.jpg

Firestrike test4 (combined), gpu usage going down - http://i.imgur.com/rtxaWWB.jpg

Firestrike normal score, note the low result in test4 - http://i.imgur.com/XBAn6ec.jpg

Farcry 4 settings page, no AA - http://i.imgur.com/hnb7RuQ.jpg

Farcry4 low gpu usage with no AA - http://i.imgur.com/DB1yrQJ.jpg

Farcry4 MSAA 4x enabled, gpu usage good - http://i.imgur.com/8zX2Ep6.jpg

GTA V settings page, no AA, everything is maxed other then AA, enchanced graphics settings are standard, did not mess with - http://i.imgur.com/mJu4EcP.jpg

GTA V with no AA, low gpu usage - http://i.imgur.com/42qOimu.jpg

GTA V with 4x AA, usage good - http://i.imgur.com/s48lnZU.jpg
 
That looks like cpu bottleneck.

If you disable AA then cpu becomes limiting factor and gpus won't work at full power.
 
Now they dont want to downclock in desktop mode. Even after restart. 1100mhz always, fans audible. So weird, it was downclocking fine and i havent changed a thing.
 
What display/displays are you running ?

144Hz @1440p won't downclock to low power state unless you cange it to use 120 Hz on desktop

Second thing might be windows power managment
 
Michaelius said:
That looks like cpu bottleneck.

If you disable AA then cpu becomes limiting factor and gpus won't work at full power.
Click to expand...

I would try using Display Driver Uninstaller and reinstalling drivers as I've read that others have had issues with low GPU usage in the same way. Maybe check Geforce.com forums. Using AA or not should not change GTA V GPU usage. Does it work if you disable SLI?
 
AriciU said:
Now they dont want to downclock in desktop mode. Even after restart. 1100mhz always, fans audible. So weird, it was downclocking fine and i havent changed a thing.
Click to expand...

Mine was doing the same thing. It was caused by a power setting in nvida control panel, changed it back to adaptive and it fixed it. What's weird is this setting is supposed to only affect clocks in 3d applications.
 
Last edited:
Killerxp100 said:
Mine was doing the same thing. It was caused by a power setting in nvida control panel, changed it back to adaptive and it fixed it. What's wired is this setting is supposed to only affect clocks in 3d applications.
Click to expand...

Reverted back to 353.06 and it's working fine again. I have it set to prefer maximum performance and it still throttles back after you close the 3d app so it should not matter how it's set. Hope it stays working.

kasakka said:
I would try using Display Driver Uninstaller and reinstalling drivers as I've read that others have had issues with low GPU usage in the same way. Maybe check Geforce.com forums. Using AA or not should not change GTA V GPU usage. Does it work if you disable SLI?
Click to expand...

With it disabled, in single card mode, it works fine everywhere. 98-99% usage always.

Michaelius said:
What display/displays are you running ?

144Hz @1440p won't downclock to low power state unless you cange it to use 120 Hz on desktop

Second thing might be windows power managment
Click to expand...

Have a 60hz display still.

SilverSliver said:
2600k at 4.7 with sli 980 Ti here. What do you need?
Click to expand...

I see that you are running 4k. Could you try setting the resolution to 1080p or 1440p in far cry 4 or gta v (if you have them available), ultra/very high settings, AA off and report your gpu usage % while in game. At 4k the problem should not appear as the gpu's are highly stressed but with lower stress on them it should behave similary to mine.

Also if you could look for gpu usage % in 3dmark firestrike normal (1080p i think it uses for normal), in the last test (combined) i'd appreciate it. Should show it in the low 60-70s.

Tnx!
 
Last edited:
Interesting, must be a bug in the hotfix driver. Hope it gets fixed soon, I find the adaptive setting causes instability on my max oc.
 
AriciU said:
I see that you are running 4k. Could you try setting the resolution to 1080p or 1440p in far cry 4 or gta v (if you have them available), ultra/very high settings, AA off and report your gpu usage % while in game. At 4k the problem should not appear as the gpu's are highly stressed but with lower stress on them it should behave similary to mine.

Also if you could look for gpu usage % in 3dmark firestrike normal (1080p i think it uses for normal), in the last test (combined) i'd appreciate it. Should show it in the low 60-70s.

Tnx!
Click to expand...

Don't have FC4 or GTA5 loaded up atm. Here is Firestrike Normal. Your 60-70% usage is correct, sometimes hitting 80s but I'd say 65-70 is pretty fair.

http://www.3dmark.com/3dm/7636804?
 
Thank you for confirming that. I can let it rest now knowing that it's not just me :)

I edited my bioses with the maxwell tweaker utility, set the vcore to 1.2v (previously one card was doing 1.187v while the other only 1.137v, no idea why, both had stock evga bios on) and increased the power limit to 350W from 275W it had by default.

Now i can set up to 140% power target in afterburner and i am stable beyond 1380 boost clk. Currently in the 1430-1443 mhz range with a 1290mhz base clock. All stable. Would crash 3dmark in seconds boosting like this with stock bios.

Best results i could get below.

FS normal - http://i.imgur.com/o8uGh9Z.jpg
FS extreme - http://i.imgur.com/08cGAnc.jpg
FS ultra - http://i.imgur.com/R6ZblRn.jpg
Valley - http://i.imgur.com/Og9x0lS.jpg (had a nasty drop in test9 i think, that's why min fps = 5)
 
Last edited:
AriciU said:
Thank you for confirming that. I can let it rest now knowing that it's not just me :)

I edited my bioses with the maxwell tweaker utility, set the vcore to 1.2v (previously one card was doing 1.187v while the other only 1.137v, no idea why, both had stock evga bios on) and increased the power limit to 350W from 275W it had by default.

Now i can set up to 140% power target in afterburner and i am stable beyond 1380 boost clk. Currently in the 1430-1443 mhz range with a 1290mhz base clock. All stable. Would crash 3dmark in seconds boosting like this with stock bios.

Best results i could get below.

FS normal - http://i.imgur.com/o8uGh9Z.jpg
FS extreme - http://i.imgur.com/08cGAnc.jpg
FS ultra - http://i.imgur.com/R6ZblRn.jpg
Valley - http://i.imgur.com/Og9x0lS.jpg (had a nasty drop in test9 i think, that's why min fps = 5)
Click to expand...

Wow, your scores are demolishing mine. I'm 1450 boost stock voltage

Here is FS ultra

http://www.3dmark.com/fs/5313786
 
Last edited:
Hmm, you're at 4.7ghz? Only 4.4 here. Must be the 4770k or I don't know why that could be.

Running win8.1 64bit, for the test i closed all background apps except afterburner and gpu-z + closed explorer.exe before starting benchmark. But that would only get a few points extra and only on the 1080 lowres setting i guess.


LE: Looking at your pic now and yeah, the graphics score is really low compared to mine. CPU should not make any diff there or does it? It should be higher for yours even since you're running it faster by 300mhz,

Something's off, doesn't make sense, test 1 181fps vs 97fps. It's like you are running only one card, wtf.


LE2: this guy is killing us both, must be running really high clocks there :) http://i.imgur.com/6W3cNpi.jpg
 
Last edited:
AriciU said:
Hmm, you're at 4.7ghz? Only 4.4 here. Must be the 4770k or I don't know why that could be.

Running win8.1 64bit, for the test i closed all background apps except afterburner and gpu-z + closed explorer.exe before starting benchmark. But that would only get a few points extra and only on the 1080 lowres setting i guess.


LE: Looking at your pic now and yeah, the graphics score is really low compared to mine. CPU should not make any diff there or does it? It should be higher for yours even since you're running it faster by 300mhz,

Something's off, doesn't make sense, test 1 181fps vs 97fps. It's like you are running only one card, wtf.


LE2: this guy is killing us both, must be running really high clocks there :) http://i.imgur.com/6W3cNpi.jpg
Click to expand...

I had a bad driver install it looks like. Ran DDU and reinstalled.

FS Ultra
http://www.3dmark.com/3dm/7639272?

I don't know why my Firestrike normal scores are so low, it's like GPU test 1 and 2 are capped at 97 fps.
 
Last edited:
Anyone try an Accelero Xtreme 3 or 4 on a 980 Ti? Can't find any results on the net.

The AX3 kept my overclocked 290X at ~35C idle and ~65C load, so I imagine results with a 980 Ti should be slightly better than any other custom cooler atm.
 
You must log in or register to reply here.
Back
Top