AMD releases Mantle, shares performance numbers

[CAT-THE-FIFTH]

http://www.fudzilla.com/home/item/33...rmance-numbers
http://www.extremetech.com/gaming/17...-cpus-not-gpus

AMD has finally released the Mantle patch and Catalyst 14.1 Beta is on the way too. However, without the new driver you won't be able to try it out in Battlefield 4. The driver release was apparently delayed because of some issues and AMD is working to sort everything out.



These issues included subpar optimization for Battlefield 4, multi-GPU support problems, intermittent stuttering when using Mantle on CrossFire systems and lack of support for some mobile technologies. More importantly perhaps, AMD also shared more info on Mantle performance gains. In Battlefield 4 we can expect a performance uplift of 2.9 to 40.9 percent depending on the scenario:

CPU-limited scenario: 40.9% (1080p) and 40.1% (1600p) performance improvement under Ultra settings and 4xAA on the AMD A10-7700K with an AMD Radeon R9 290X.
GPU-limited scenario: 2.7% (1080p) and 1.4% (1600p) performance improvement under Ultra settings and FXAA on the Core i7-4960X with an AMD Radeon R7 260X
Average uplift for 1080p: 13.28% (Average of 290X and 260X data on the i7-4960X, A10-7700K, FX 8350 and i5-4670K)
Average uplift for 1600p: 11.35% (Average of 290X and 260X data on the i7-4960X, A10-7700K, FX 8350 and i5-4670K)

StarSwarm from Oxide Games promises even more impressive gains:

CPU-limited scenario: 319% (1080p) and 281% (1600p) performance improvement in the “RTS” test on Extreme settings with the AMD A10-7700K and an AMD Radeon R9 290X.
GPU-limited scenario: 5.1% (1080p) and 16.7% (1600p) performance improvement in the “RTS” test on Extreme settings with the Core i7-4960X and an AMD Radeon R7 260X
Average uplift for 1080p: 115.65% (Average of 290X and 260X data on the i7-4960X, A10-7700K, FX 8350 and i5-4670K)
Average uplift for 1600p: 75.19% (Average of 290X and 260X data on the i7-4960X, A10-7700K, FX 8350 and i5-4670K)

We should see the first independent reviews in a matter of days, as soon as the Catalyst 14.1 issues are resolved, but AMD's numbers are very good indeed. The biggest gains come in CPU limited scenarios, as Mantle is supposed to cut down plenty of processor overhead associated with DirectX.

Like we said, Mantle is not just about GPUs, in fact it might have a bigger impact in the APU department - so it's more of shot at Intel than Nvidia. Catalyst 14.1 also happens to be AMD's first HSA enabled driver, but so far only a handful of applications stand to benefit from HSA. AMD points out that Mantle makes less of an impact on high resolutions with max detail settings, as they are GPU-limited scenarios.

However, in CPU limited scenarios it starts to make a lot more sense. In other words you'll get higher performance gains with a relatively weak APU paired with a powerful discrete GPU. The average numbers are quite a bit lower, but they're still in double digit territory in BF4 and they are much higher in StarSwarm.

The numbers are promising and we'll try to put AMD's claims to the test soon on a Kaveri - Hawaii combo.

[shaithis]

Now we finally see why AMD has pushed it so hard and nVidia haven't bothered. It's only really a benefit for weaker CPUs and therefore will make their APUs look considerably better. Bit of a let-down really.

[kalniel]

Now we finally see why AMD has pushed it so hard and nVidia haven't bothered. It's only really a benefit for weaker CPUs and therefore will make their APUs look considerably better.

That's what we thought and what AMD said all along - mantle reduces the CPU overheads/few-threaded limitations from DirectX. This info isn't new, we've just got some performance figures to back it up.

[CAT-THE-FIFTH]

It became more and more obvious it was for lower end CPUs,which I even mentioned on here and OcUK for ages. Most of the AMD slides were with Kaveri and massively underclocked FX CPUs. With AMD APUs being over 75% of their sales,its no wonder they wanted to make their APUs more competitive,but it also means their FX CPUs will probably get a reasonable performance bump.

It might be clearer now why the A10 Kaveri has 512 shaders,even though bandwidth is an issue - in Mantle enabled games the uplift might be noticeable with both the IGP and in dual graphics scenarios which would help them fight Intel better.

Nvidia really does need to do something,since most sales are of low end CPUs and graphics cards,and most gaming systems are not Core i5 K series CPUs with £200+ graphics cards. If AMD gets a decent bump in IGP and low end card performance with this,together with the big gains Intel is making in IGP performance,it is a threat to Nvidia.

Once we get the GTX750TI reviews it will be an interesting preview of what Maxwell brings to the table. Nvidia mentioned some serious efficiency improvements with it a while ago. Things might get very interesting this year.

Edit!!

A bit of side news:

http://www.pcper.com/reviews/Graphic...ng-Phase-2-HSA

That changes today as Phase 2 of the AMD Frame Pacing feature has finally been implemented for products that do not feature the XDMA technology (found in Hawaii GPUs for example). That includes HD 7000-series GPUs, the R9 280X and 270X cards, as well as older generation products and Dual Graphics hardware combinations such as the new Kaveri APU and R7 250. I have already tested Kaveri and the R7 250 in fact, and you can read about its scaling and experience improvements right here. That means that users of the HD 7970, R9 280X, etc., as well as those of you with HD 7990 dual-GPU cards, will finally be able to utilize the power of both GPUs in your system with 4K displays and Eyefinity configurations!

[CAT-THE-FIFTH]

Some figures from DICE:

http://battlelog.battlefield.com/bf4...tle-live/last/
http://wccftech.com/dice-officially-...4-123-gb-size/

[CAT-THE-FIFTH]

http://www.neogaf.com/forum/showpost...0&postcount=28



A10 7805K IGP and BF4



FX8350 with HD7970 and BF4



Core i7 3970X with R9 290X Crossfire and BF4