• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

290 CF thoughts/impressions

n0x1ous

n0x1ous

Platinum Member
Finally got some time to play around with the new (to me anyway) cards last night. Power consumption is just incredible. Playing BF4 multi-full 64p server, Peak power pulled from the wall was 800 watts averaging around 750 (measured using Kill-a-watt meter).

With the same system and 7970 CF it rarely crossed 600 watts. The FX doesnt have the balls to push a consistent 120hz on such a CPU intensive game, so I tried Metro LL to let it stretch its legs.

Turned everything to max, (except no PhysX obviously) and never dropped below 50 fps. Averaging 60-70. Mind you, that is with 4x SSAA with is basically rendering at 4k and downsampling to 1080p. The image quality provided by that was amazing. XDMA crossfire is super-smooth and both cards were pegged at 100% usage.

All in all, Hawaii is a very impressive chip, and I hope AMD continues to bring out large die chips to compete with Nvidias big dogs.
 
Last edited:
You should also test the New Tomb Raider with 4xSSAA. To me this looked substantially better than FXAA. Should be able to lock to 60+ on that one.

I've been messing around with a single PCS+290 with lowered clocks/volts/power target. I'm considering slamming 2 of em into a htpc that has successfully housed 7950 xfire.

This card defaults with a +50mv increase and stock clocks are 1040/1350. This is very good performance and close to stock 290x level. 11300 in graphics score for Firestrike.

I can get 88% of that performance with:
*900c
*1250m
*-50mv
*-30%power target

Going to 950c requires more tinkering on my part for undervolt, but it nets 92% of stock (1040/1350) performance.

I have not gotten a chance to use a KillaWatt meter to check for power use drop, but based on temps I believe those settings (900c) saves 20-25% power. So for crossfire it'd be potentially close to 7970 level of power use while still blowing the performance away.
 
Last edited:
Face2Face said:
Did you enable Mantle? Also, what models did you purchase?
Click to expand...

Yes it is with Mantle enabled. Its way better than DX11 in FPS, frame time, and smoothness.

With a single 290 @ 1050 core in Mantle I'm usually getting 70-80 FPS (this is 1080p all Ultra no SS)

With crossfire @ stock clocks, I get 85-100.

My 780 SLI intel system bounces off of 120 consistently.
Difference? FX can't do it even @ 4.8. 3770K can do it easily @ 4.5.

I know if I put the radeons in the Intel system it would do 120 easily, but I don't want to go through with the hassle at this point. I bought the FX because I had never owned an AMD CPU before and wanted to experience for myself how it was. Its a decent chip for most things (most games are GPU limited and its good for premier rendering etc etc) but it can't keep up in CPU intensive games like Crysis 3 or Bf4 even with Mantle.

Despite that, I still prefer to play BF4 on the AMD system because for me the smoothness and low input lag of Mantle feels better than the extra frames.

290's are reference cards from ebay - sold my 7970's for pretty decent amount and only had to spend a few bucks to upgrade.
 
Last edited:
I ran CF 290 for a while; it worked fine for BF3 (not BF4, I don't play military FPSes much so I'm way behind), Tomb Raider 2013, L4D2, etc. just fine. Even at 5760x1080, BF3 cranked to near-max on CF 290 felt fairly smooth.

Problem: the pair got hot despite my very well-ventilated case with tons of fans. I had either had to crank up the fan or else undervolt to get the top card to stop throttling at stock speeds. Granted it was a ref card, but open-air cards wouldn't do much better if at all, since they would be sucking up hot air from the lower card as well. Ultimately I broke up the pair by giving the second card to my gf. I retained the other card, an ASUS DirectCUII.
 
blastingcap said:
I ran CF 290 for a while; it worked fine for BF3 (not BF4, I don't play military FPSes much so I'm way behind), Tomb Raider 2013, L4D2, etc. just fine. Even at 5760x1080, BF3 cranked to near-max on CF 290 felt fairly smooth.

Problem: the pair got hot despite my very well-ventilated case with tons of fans. I had either had to crank up the fan or else undervolt to get the top card to stop throttling at stock speeds. Granted it was a ref card, but open-air cards wouldn't do much better if at all, since they would be sucking up hot air from the lower card as well. Ultimately I broke up the pair by giving the second card to my gf. I retained the other card, an ASUS DirectCUII.
Click to expand...

Yes - its taking about 60% fan on the top card to prevent throttling. I game with headphones most of the time so noise isn't THAT big of deal to me, but I can hear them even with headphones on. 780 SLI I cannot hear with headphones on. The 290s are not bad enough to make me change anything yet, but if it bugs me enough then ill throw a couple CLC's on them.
 
n0x1ous: Good post and observation.

I run 2 GTX 670FTWs in SLI in my 3770k @ 4.5Ghz rig and like that but just got a Sapphire R9-290 Tri-X for my FX8350 @ 4.6Ghz rig. Your power draw observations are interesting. My EVGA GTX780 Classified is OC'd to 1306 on the core and is my 3930k @ 4.6Ghz rig.

I had replaced a GTX680 in the FX8350 rig and the R9 290 is noticeably faster. I can only imagine what 2 in CF would be like.

Since I have single monitors for each of my rigs, Achieva Shimian 27" 2560 x1440 for the 3930k, Acer 27" 1920 x1080 for the 3770k and Hanns G 27.5" 1920 X 1200 for the FX8350, I have plenty of horsepower to run games nicely.

Are you running single monitors on the GTX 780 SLI and R9-290 CF rigs?
 
guskline said:
n0x1ous: Good post and observation.

I run 2 GTX 670FTWs in SLI in my 3770k @ 4.5Ghz rig and like that but just got a Sapphire R9-290 Tri-X for my FX8350 @ 4.6Ghz rig. Your power draw observations are interesting. My EVGA GTX780 Classified is OC'd to 1306 on the core and is my 3930k @ 4.6Ghz rig.

I had replaced a GTX680 in the FX8350 rig and the R9 290 is noticeably faster. I can only imagine what 2 in CF would be like.

Since I have single monitors for each of my rigs, Achieva Shimian 27" 2560 x1440 for the 3930k, Acer 27" 1920 x1080 for the 3770k and Hanns G 27.5" 1920 X 1200 for the FX8350, I have plenty of horsepower to run games nicely.

Are you running single monitors on the GTX 780 SLI and R9-290 CF rigs?
Click to expand...

Yes - single Asus VG248QE so all that extra horsepower for 1080p is for 120hz. Planning on 4k whenever I see a realistic mix of IPS, price, Gsync/Freesync.

I am curious what kind of FPS you are getting in Mantle on your Tri-X? FWIW, power draw in BF4 with 1 - 290 was around 550-600 (again total system power from the wall) watts. My FX is @ 1.45 V with heavy LLC

Thats a nice stable of rigs you have there. which is your favorite?
 
Last edited:
Attic said:
You should also test the New Tomb Raider with 4xSSAA. To me this looked substantially better than FXAA. Should be able to lock to 60+ on that one.
Click to expand...

I do have this, but never started as its in a long backlog. Will definitely try it @ 4XSSAA!
 
n0x1ous said:
Yes - single Asus VG248QE so all that extra horsepower for 1080p is for 120hz. Planning on 4k whenever I see a realistic mix of IPS, price, Gsync/Freesync.

I am curious what kind of FPS you are getting in Mantle on your Tri-X? FWIW, power draw in BF4 with 1 - 290 was around 550-600 (again total system power from the wall) watts. My FX is @ 1.45 V with heavy LLC

Thats a nice stable of rigs you have there. which is your favorite?
Click to expand...

Okay, my Kill A Watt meter measured @525watts (low as 480 never exceeded 535W) for my FX 8350 system without the monitor.

I'm using a Asus Sabertooth FX990 Rev1 mb with the FX8350 set to 23 multiplier by 200 fsb. Vcore is set at 1.468 manually. I have the cpu custom water cooled with a XSPC EX 360 rad, Swiftech 655-b pump in a XSPC Acrylic D5 bay/res with a EK Supreme LTX AMD CSQ waterblock.

The Sapphire R9 290 Tri-X is Overclocked to 1060 core and 1375 memory.

BF4 multiplayer on my Hanns G 1920 x1200 monitor was showing 88 to 125 fps in Mantle with most gameplay @105 fps.

My 3770k is in an Asus P8Z68-V Pro/Gen3 with the cpu oc'd to 4.5Ghz and 2 EVGA GTX670 FTWs in SLI. A fast devil.

Hard to choose what is my favorite but I have the most $$ tied up in my 3930k rig with the specs below
 
Last edited:
Just out of curiosity, does anyone have Planetside 2 installed, and can it be run at 4k with a single 290 with some settings turned down? 😀
 
guskline said:
Okay, my Kill A Watt meter measured @525watts (low as 480 never exceeded 535W) for my FX 8350 system without the monitor.

I'm using a Asus Sabertooth FX990 Rev1 mb with the FX8350 set to 23 multiplier by 200 fsb. Vcore is set at 1.468 manually. I have the cpu custom water cooled with a XSPC EX 360 rad, Swiftech 655-b pump in a XSPC Acrylic D5 bay/res with a EK Supreme LTX AMD CSQ waterblock.

The Sapphire R9 290 Tri-X is Overclocked to 1060 core and 1375 memory.

BF4 multiplayer on my Hanns G 1920 x1200 monitor was showing 88 to 125 fps in Mantle with most gameplay @105 fps.

My 3770k is in an Asus P8Z68-V Pro/Gen3 with the cpu oc'd to 4.5Ghz and 2 EVGA GTX670 FTWs in SLI. A fast devil.

Hard to choose what is my favorite but I have the most $$ tied up in my 3930k rig with the specs below
Click to expand...

Interesting.... You probably have lower CPU temps and gpu temps which would explain the slightly lower power use. I also cap bf4 to 120fps via config file. I can get to 120 in a few spots or if I look up at the sky..... the fps numbers I gave earlier were from Guillen Peaks FWIW
 
Last edited:
n0x1ous said:
Interesting.... You probably have lower CPU temps and gpu temps which would explain the slightly lower power use. I also cap bf4 to 120fps via config file. I can get to 120 in a few spots or if I look up at the sky..... the fps numbers I gave earlier were from Guillen Peaks FWIW
Click to expand...
Just played @20 minutes of BF4 multiplayer (48) and recorded cpu and gpu temps and vcore via Hardware Pro. CPU package was MAX 50C with FX8350 temp showing 47.3C and vcore max of 1.464.

GPU temp max was 74C (gotta love that Sapphire Tri-X cooler even OC'd to 1060 core 1375 memory!)
 
guskline said:
Just played @20 minutes of BF4 multiplayer (48) and recorded cpu and gpu temps and vcore via Hardware Pro. CPU package was MAX 50C with FX8350 temp showing 47.3C and vcore max of 1.464.

GPU temp max was 74C (gotta love that Sapphire Tri-X cooler even OC'd to 1060 core 1375 memory!)
Click to expand...

Yeah my reffy cooled cards are usually 92-93C
CPU is about the same as yours....
 
I played some more BF4 multiplayer last night with crossfire on and the FX just cannot handle it. lots of stutters and cpu spikes that arent there when using single GPU. Im debating swapping out the CHvF-Z and 8350 for a Z97.

But then I think, its just 1 or 2 games so why bother?.....
 
Great post. BF4 is super CPU intensive. You can knock geometric detail down one notch, and you should gain a decent amount of FPS. It has a big impact on CPU performance. Or it does for me anyway.
 
Stuka87 said:
Great post. BF4 is super CPU intensive. You can knock geometric detail down one notch, and you should gain a decent amount of FPS. It has a big impact on CPU performance. Or it does for me anyway.
Click to expand...

Right now I just turn crossfire for BF4 off since it barely improves frame rates at all, but adds stutter due to CPU limitation.

Thanks for suggestion, but not interested in turning down any details. I also have my intel/nvidia rig which can play at locked 120hz at max detail to fall back on.

The AMD system was more of an experiment and a chance to play with Mantle....And my experience is AMD has great GPU's and Mantle is awesome, but FX has major limitations even in a well threaded game like BF4

http://www.anandtech.com/show/6934/choosing-a-gaming-cpu-single-multigpu-at-1440p/7
I had read this article before but it was a good refresher....In most games that are GPU limited, FX keeps up fine for multi-gpu, but on CPU intensive games like Civ 5, BF4, Crysis 3 it gets crushed and cannot scale.
 
Last edited:
I think the FX 8350 when OC'd to this level must be drawing 175-200+watts.

I'm messing around with 290 xfire and power draw, mainly in effort to get noise way way way down. They can be much more efficient that rep has got them.

It seems that the cards get really nasty with power draw when OC'd and overvolted and are somewhat nasty w/ power draw w/ ref coolers when temps are allowed to exceed 84c-85c.

You may want to test your power draw if you set fan speed to 100% and see if temps remain below 84c and what that does for power draw.
 
N0xious have you checked for core parking......that can cause hitching in BF4; I know it was with me running mantle on my 8120 with my 7870 LE........and I also have 120 monitor........
 
I would assume you guys saw the [h] 290x crossfire vs. 780 SLI review. The 290's are amazing cards and ironically even smoother in crossfire than the 780 is in SLI.
 
ok; what's your OC on your system NB? sometimes not right I've seen other FX systems with 290s crossfire handle BF4 without issues. what card did you have in before; did you ddu the drivers? 🙂 ram speed; lots can be effecting.....
 
MeldarthX said:
ok; what's your OC on your system NB? sometimes not right I've seen other FX systems with 290s crossfire handle BF4 without issues. what card did you have in before; did you ddu the drivers? 🙂 ram speed; lots can be effecting.....
Click to expand...

2600 on NB and HT. Ram speed is 1866

previous cards were 7970's.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top