BF3 patch includes 7xxx series perf improvements

[Skurge]

http://battlelog.battlefield.com/bf3/news/view/2832654778817797165/

The new patch says it has some improvements for Radeon 7xxx series. I don't have my 7950 yet, so I can't test anything. If anyone has the game and hasn't patched it yet, can you let us know if performance has gone up.

Lets hope this is a result of AMD starting to work more closely with devs.

 

[Stuka87]

I won't be able to patch until tonight.

Is anybody else noticing that maps take ages to load first time you join a game? I remember in BC2 there was a fix from AMD that fixed map loading issues.

 

[blackened23]

I won't be able to patch until tonight.

Is anybody else noticing that maps take ages to load first time you join a game? I remember in BC2 there was a fix from AMD that fixed map loading issues.

First load is slow on all hardware.

 

[kingsownall]

I did not see this patch. Curious to see the gain in performance.

 

[RavenSEAL]

I heard it's wrecking punkbuster. Nerds are raging.

 

[ss284]

I haven't gotten home to patch yet, but I am a nerd, and I will probably be raging tonight.

 

[djnsmith7]

It patched this morning as I was trying to do something else, so I didn't pay much attention to it. I'll check it out later.

 

[boxleitnerb]

According to PCGH this patch will only improve FXAA performance, not MSAA.

 

[blackened23]

I just tried it on 64 player conquest and my framerates were ridiculous, triple digits @ 2560x1440 with fxaa and triple digits with 4x msaa as well, but hovered around 85-90 in dense city areas with a lot of activity. Both GPU's pegged at 99% usage the entire time. GPUs clocked at 1150/1700x2.

I have no basis of comparison though, it "feels" a bit faster than before. I'm eagerly awaiting the AdamK report

 

[superjim]

I just tried it on 64 player conquest and my framerates were ridiculous, triple digits @ 2560x1440 with fxaa and triple digits with 4x msaa as well, but hovered around 85-90 in dense city areas with a lot of activity. Both GPU's pegged at 99% usage the entire time. GPUs clocked at 1150/1700x2.

I have no basis of comparison though, it "feels" a bit faster than before.

I remember reading you picked up at least one 7970 in another thread, sounds like you have them in CF. Update your sig, it confuses me :sneaky:

 

[blackened23]

The performance increase from this patch is pretty substantial. I thought I was imagining it, but many users at H are reeporting it as well.

:thumbsup::thumbsup::thumbsup:

 

[lavaheadache]

All they did was take the line of code out that handicrapped amd cards intentionally, lol.

 

[blackened23]

I have no documented or screenshot basis for comparison though, I hope someone does a before and after type of thing (AdamK?) :sneaky:

 

[djnsmith7]

I didn't have time to take any screenshots or benchmarks before the patch, so I can't say for sure, but my sentiments are in line with those saying the framerates increased. I won't say I'm impressed with the numbers, but I'm hoping more efficient (mature) drivers will increase the Min & Avg figures.

My FRAPS numbers after the patch:

1x 7970 (1125 / 1575):

Min - 47
Avg - 110.7
Max - 186

2x 7970 (1125 / 1575):

Min - 61
Avg - 135.6
Max - 201

 

[RavenSEAL]

I didn't have time to take any screenshots or benchmarks before the patch, so I can't say for sure, but my sentiments are in line with those saying the framerates increased. I won't say I'm impressed with the numbers, but I'm hoping more efficient (mature) drivers will increase the Min & Avg figures.

My FRAPS numbers after the patch:

1x 7970 (1125 / 1575):

Min - 47
Avg - 110.7
Max - 186

2x 7970 (1125 / 1575):

Min - 61
Avg - 135.6
Max - 201

Damn, that's barely 50% scaling D:

 

[blackened23]

Hmm, that doesn't seem right - are you using the RC11 driver? My scaling was substantially more than that. I'm running 2560x1440 though. I definitely get near 90%-100% scaling, both of my GPU's are constantly pegged at 99% in the MSI OSD overlay.

My scaling is similar to what i've seen here:

http://www.hardwareheaven.com/revie...sfire-graphics-card-review-battlefield-3.html

What are your GPU usage numbers?

 

[TakeNoPrisoners]

Damn, that's barely 50% scaling D:

I'd go out on a limb and say that is a CPU bottleneck.

 

[djnsmith7]

Hmm, that doesn't seem right - are you using the RC11 driver? My scaling was substantially more than that. I'm running 2560x1440 though. I definitely get near 90%-100% scaling, both of my GPU's are constantly pegged at 99% in the MSI OSD overlay.

My scaling is similar to what i've seen here:

http://www.hardwareheaven.com/revie...sfire-graphics-card-review-battlefield-3.html

What are your GPU usage numbers?

For some reason, MSI OSD has them ranging from 70 - 80%. Not sure why. I thought my 1kw PSU might have been the culprit, but I just put the 1500w in & the numbers haven't moved. How do I get these bad boys in the high 90s?

Play @ 1080p.

Update:

Ran FRAPS again & here are the new numbers with the new PSU (previously only gauged GPU usage %). Numbers improved, but not by much. GPU usage % still hovering in the 70 - 80% usage range for each card.

Min - 72
Avg - 140
Max - 201

 

[djnsmith7]

I'd go out on a limb and say that is a CPU bottleneck.

A 2500k @ 4.5 a bottleneck? Not likely.

 

[videogames101]

A 2500k @ 4.5 a bottleneck? Not likely.

Check BF3 benches before you say that, 64-player is a CPU monster.

 

[RavenSEAL]

I'd go out on a limb and say that is a CPU bottleneck.

I mean, the guy is already racking up 110FPS AVG. So it's not like it's a problem.

However, I still don't understand the idea behind xFire 7970s at this point, one card alone wrecks anything out there.

 

[UJ09]

I don't know why people keep saying that 4Ghz is enough for crossfire for BF3. I'm running 2 X 7970 crossfire at 4Ghz(2500k), 1080p max with only FXAA(no MSAA yet), OSD shows that I only managed to run both cards around 50%-60% most of the time, and it's stuttering and even at 60fps it feels like playing at 24fps, really annoying.

I switched back to single card mode, it's so smooth and enjoyable, comparing single card and crossfire the different is like night and day.

Try most of the drivers, cap or without cap pretty much the same result. So either CPU is the bottleneck factor or something wrong with my system or driver, but I think CPU is the main factor here.

 

[Olikan]

All they did was take the line of code out that handicrapped amd cards intentionally, lol.

meh, i don't belive this.

the 7XXX series is very diferent than the old cards...

IMO, it' just a code, that work well with Vliw, but not with GCN

 

[Arkadrel]

Ran FRAPS again & here are the new numbers with the new PSU (previously only gauged GPU usage %). Numbers improved, but not by much. GPU usage % still hovering in the 70 - 80% usage range for each card.

Min - 72
Avg - 140
Max - 201

70% GPU usage and 140 FPS avg not bad.

 

[blackened23]

A 2500k @ 4.5 a bottleneck? Not likely.

I'd look into a solution to this, everyone I know running a similar setup has 90-99% GPU usage 100% of the time when vsync is turned off. You should definitely be getting upper 90% gpu usage in bf3. Something is awry, trust me. I have a 2600k @ 4.8 so cpu probably isn't a bottleneck. You're not getting throttled are you? If you pull up gpu clock numbers in the overlay, if they fluctuate it could be a powertune issue (not being high enough), i've seen low gpu usage in cases like that with heavy overclocks..(due to throttling)

I'm using the RC11 driver with the latest CAP dated 2/15 on the AMD website