• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

2 x GTX 670 @ 1440p good enough for BF4 Ultra?

Associate
Joined
12 Nov 2011
Posts
89
Hi Guys,

I am looking to upgrade shortly for BF4. I was in two minds getting a 3 monitor setup, but frankly, I think the cost involved is to much as I would like to run the game in Ultra Settings.

I will likely play the game on a 27inch 1440p. Now I know there are no official specs for the game yet, but would 2 x 670 2GB cards handle this resolution and likely run the game on Ultra?

Reason being, these cards are coming down in price and would ideally like them to last the next 2-3 years really before I do a brand new rebuild.

Many Thanks
 
Grunt wise you should be fine, vram wise you will be lacking. No caching present in these alpha benchmarks.

qed2yJa.jpg
 
Ok well I can easily get the 4GB thats no problem. Do you think it will run on Ultra.
My current Spec is only PCIe2.0 i7 2600k OC @4.8GHz 16GB DDR3 RAM
 
Ok well I can easily get the 4GB thats no problem. Do you think it will run on Ultra.
My current Spec is only PCIe2.0 i7 2600k OC @4.8GHz 16GB DDR3 RAM

If you can run 1440p ultra with X4AA nicely now then you should be ok for Battlefield 4 i reckon. I doubt it will be that much more demanding than Battlefield 3, maybe we'll have to knock AA down to x2 or something.
 
I dont have a 1400p at the moment. My current card is GTX 580 Lightning 1.5GB
I currently have 1050p 120Hz Monitor at the moment.
 
I dont have a 1400p at the moment. My current card is GTX 580 Lightning 1.5GB
I currently have 1050p 120Hz Monitor at the moment.

I imagine a 670 SLI setup can power Battlefield 3 at 1440p with all the bells and whistles on without breaking sweat. Maybe someone with the setup here will confirm.
 
670 SLi kept my 1440p at a constant 60fps with everything cranked up on BF3.

I doubt Battlefield 4 will tax 670 sli or any of the next gen games coming this year tbh.
 
670 SLi kept my 1440p at a constant 60fps with everything cranked up on BF3.

I doubt Battlefield 4 will tax 670 sli or any of the next gen games coming this year tbh.
I suspect this as well, currently using sli 670's at 1920x1200, overkill tbh as one ran bf3 with ease. But id like to move to a larger monitor or a 120Hz model.
 
The results from alpha near the top are just that alpha results. There would have been no optimizing yet done so hopefully when optimized for release the usage might come down a canny bit. Cos personally it looks like bf3 to me so i really cant see it really needing all that much when bf3 would run fine on a 2gb card or 1.280mb card like my old 570.

Best to wait and see when game comes out before upgrading, plus prices might be cheaper then.
 
Matt you always roll out that chart when the AA VRAM rears its head, but both of those cards in the test are 3GB cards and we know from various people's experience on here that the monitoring programs show that the cards like to gobble up available VRAM even if it is not used by the game. So really the only fair conclusion would be if the test was run with 2GB 680s/670s to see if the buffer is reached and fps drops significantly.
 
Yeah ive noticed that too. If you have a 1280mb card it would show less. If you have a 2gb card it would show more is used only because u have the space but wouldnt matter really. If you had a 4gb card maybe it would use more.

Games and vram amounts scale as needed. Dont they? So if you use a 1280mb card it might use 1gb for the game but still performs great. If you have 2gb card it might use 1.5gb vram and still runs great just like it would with the 1280mb card. Having less doesnt really mean it performs less. Tho if under 1gb cards they could as that is getting a bit low these days like say a 460 768mb card.
 
Yeah ive noticed that too. If you have a 1280mb card it would show less. If you have a 2gb card it would show more is used only because u have the space but wouldnt matter really. If you had a 4gb card maybe it would use more.

Games and vram amounts scale as needed. Dont they? So if you use a 1280mb card it might use 1gb for the game but still performs great. If you have 2gb card it might use 1.5gb vram and still runs great just like it would with the 1280mb card. Having less doesnt really mean it performs less. Tho if under 1gb cards they could as that is getting a bit low these days like say a 460 768mb card.

I laughed my nuts off the other day when my HD 5970 ran the Tomb Raider benchmark fully maxed with no problems, it is a 1gb per GPU card lol.

Single HD 5970 running the game fully maxed out @1080p

ngfu.jpg

Not bad for 1gb of vram per GPU.:D
 
The Frostbite engine will use what you have if it needs too. Showing near to max on a 2gb card just means the game is using what is available.

Tbh the whole VRAM thing is a big panic over nothing. Maybe with multi screens or ultra high res setups there may be issues but I have not seen anything running up to 2560x1440 in game or even in my brief experience with triple 1080p screens.

I would bet you would run out of GPU grunt long before VRAM becomes a major concern.
 
Grunt wise you should be fine, vram wise you will be lacking. No caching present in these alpha benchmarks.

qed2yJa.jpg

im glad i invested in 4gb cards now, I hope there's some optimizations to come in the game and on the driver side. Looking at the 690 scores ill won't be getting 60fps on my setup :(
 
im glad i invested in 4gb cards now, I hope there's some optimizations to come in the game and on the driver side. Looking at the 690 scores ill won't be getting 60fps on my setup :(

I will back my two GTX 690s to beat anyone else's two cards on a single screen running BF4.:D

This could be unfortunate for me if someone here buys a pair of HD 7990s.:D
 
How can other maps be cached when the alpha was only 2gb in size and contained only one map?

I don't think it works like that. Unless you have evidence to the contrary the fact only 1 map is being played would seem to make sod all difference to what the card chooses to cache.
 
Back
Top Bottom