AMD Mantle API [updating]

[neliz]

If I only had time to test I would do it properly
These were just static scene frame rates and they fluctuate +/- 5FPS hence 'average' word was used. But because it was static scene you can calculate frame time for this specific spot.

Maybe tomorrow I will have a bit of time to test Mantle more thoroughly. For sure I have issues with Mantle and CF on my setup as this I managed to try already and it just crashes within few seconds.

you have the 290X in CrossFire? BTW, hope you're happy with the 290X GAMING

 

[Malo]

People buy 290X for GAMING??

 

[Lightman]

you have the 290X in CrossFire? BTW, hope you're happy with the 290X GAMING

Yes, but mostly CF is disabled as I game on one card while the other is mining.
I also have 2xR9 290 in my other PC (HTPC) of which one is excellent MSI TwinFrozr 4G. Super quiet and based on Hynix memory

ps. Are you still playing BF3(4)?

 

[itsmydamnation]

my 2600@4.3 kick the butt of your i920 "real i7"

whooptydo.... that has nothing to do with my point..... total system costs especially on LGA 1366 where significantly different to 1155(?) and thus a 860/2600/3770/4770 provide a good upperbound on what most serious gamers would/will buy.

 

[Kaotik]

Hmm.. so it looks like Mantle performance does not improve much with CPU clockspeed? From 2Ghz onwards it did not gain much, as long as the CPU has/is using 4 cores. Or could that be a GPU bottleneck? But from what I could gather from German it is only 1080p so probably not?

Or then yes, Oxide said that in their tests they could downclock 8-core FX to 2GHz without it becoming a bottleneck for R9 290X

 

[neliz]

People buy 290X for GAMING??

Silly me

Yes, but mostly CF is disabled as I game on one card while the other is mining.
I also have 2xR9 290 in my other PC (HTPC) of which one is excellent MSI TwinFrozr 4G. Super quiet and based on Hynix memory

ps. Are you still playing BF3(4)?

I'm curious about your temps etc. Can you PM them to me?

occasionally, but completely not like last year.

 

[elect]

I read a guy mentioning that his 290x was also about 5° cooler under Mantle

Some of you have heard something related to heat as well?

 

[Gipsel]

I read a guy mentioning that his 290x was also about 5° cooler under Mantle

Some of you have heard something related to heat as well?

If a system was significantly bottlenecked by the CPU and the framerate rises with Mantle, I would expect a temperature rise to be more likely as the GPU does more now.

 

[elect]

If a system was significantly bottlenecked by the CPU and the framerate rises with Mantle, I would expect a temperature rise to be more likely as the GPU does more now.

I'd too, but this is what I read

 

[Nemo]

http://www.pcgameshardware.de/AMD-R...ecials/AMD-Mantle-Test-Battlefield-4-1107754/

 

[Kyyla]

http://www.pcgameshardware.de/AMD-R...ecials/AMD-Mantle-Test-Battlefield-4-1107754/

That is super impressive stuff. I couldn't be happier. Also my 290 finally arrived in the mail

 

[UniversalTruth]

That is super impressive stuff.

Except that on both Mantle images, there is some very strange kind of fog which is not present on DX... hmmm

 

[Dave Baumann]

Which is also addressed in the article (and this thread) via comments from Repi.

 

[DieH@rd]

New Star Swarm Build: Better Mantle Performance
http://oxidegames.com/2014/02/04/new-star-swarm-build-better-mantle-performance/


We just deployed a new build of our Star Swarm stress test that significantly improves the demo’s performance using AMD’s Mantle API. We suggest that those of you interested in benchmarking and performance numbers re-run your Mantle scenarios; you may be surprised at the results.

Did we crack some secret code or find a crazy new optimization? No, nothing so spectacular. The truth is that we made a mistake in our haste to deploy the build that stripped out the activation process. We didn’t follow our normal release process, and missed how a minor change in that build had disabled some of the Nitrous engine’s multi-threading features when using Mantle. Unfortunately we didn’t notice that at first, as nobody was running Mantle last week due to the beta driver being delayed.

It’s all fixed now, so Mantle users should see a noticeable boost in performance on most configurations (this fix doesn’t have a huge effect on powerful-CPU/slow-GPU systems). We’ve also addressed the “gray screen” issue that was affecting some non-English Windows systems.

We appreciate everyone’s patience and feedback as we hunt down some of these hard-to-reach bugs for this alpha test of the Nitrous Engine.





Ok, people. It's time for new bench numbers!

 

[sebbbi]

Nitrous says 30k batches at 30 fps (DirectX + high end GPU). How many can the Mantle version of Star Swarm push?

(I am not impressed if it's not more than 250k at 60 fps)

 

[mc6809e]

http://www.pcgameshardware.de/AMD-R...ecials/AMD-Mantle-Test-Battlefield-4-1107754/

The core scaling chart is very interesting, especially the jump from two cores+SMT to four cores+SMT.

D3D11 performance 67fps / 42fps = 1.60
Mantle performance 125fps / 63fps = 1.98

So it seems at least some of Mantle's performance can be attributed to increased concurrency rather that just shorter execution times due to a simplified and more efficient API.

 

[Andrew Lauritzen]

(I am not impressed if it's not more than 250k at 60 fps)

From a brief look at the demo yesterday it would become GPU/vertex limited before that, as it ultimately isn't *only* a batch benchmark . Seems to get limited around ~100k depending on your GPU. In the earlier talks they mention that when they get GPU limited they are still only using ~half a quad core CPU or so, although obviously measurements like that (i.e. w/ idle) are never great due to power, turbo, etc.

It's easy enough to play around with though. All the settings just live in some INI files with fairly obvious names.

Regardless, for such benchmarks you always have to define what states are changing between the 250k calls If you don't change any, you can already get that in DX of course.

 

[xDxD]

New Star Swarm Build: Better Mantle Performance
http://oxidegames.com/2014/02/04/new-star-swarm-build-better-mantle-performance/


We just deployed a new build of our Star Swarm stress test that significantly improves the demo’s performance using AMD’s Mantle API. We suggest that those of you interested in benchmarking and performance numbers re-run your Mantle scenarios; you may be surprised at the results.

Did we crack some secret code or find a crazy new optimization? No, nothing so spectacular. The truth is that we made a mistake in our haste to deploy the build that stripped out the activation process. We didn’t follow our normal release process, and missed how a minor change in that build had disabled some of the Nitrous engine’s multi-threading features when using Mantle. Unfortunately we didn’t notice that at first, as nobody was running Mantle last week due to the beta driver being delayed.

It’s all fixed now, so Mantle users should see a noticeable boost in performance on most configurations (this fix doesn’t have a huge effect on powerful-CPU/slow-GPU systems). We’ve also addressed the “gray screen” issue that was affecting some non-English Windows systems.

We appreciate everyone’s patience and feedback as we hunt down some of these hard-to-reach bugs for this alpha test of the Nitrous Engine.





Ok, people. It's time for new bench numbers!

how can a TechDemo be significant?

 

[DSC]

http://techreport.com/review/25995/first-look-amd-mantle-cpu-performance-in-battlefield-4

Heck, on the 4770K, the GTX 780 Ti with D3D outperforms the R9 290X with Mantle.

 

[Picao84]

http://techreport.com/review/25995/first-look-amd-mantle-cpu-performance-in-battlefield-4

I think that was more or less known on Single Player? It looks like Mantle helps more on Multiplayer where CPU is a bigger bottleneck?