The AMD Mantle Thread

Page 197 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Status
Not open for further replies.

bystander36

Diamond Member
Apr 1, 2013
5,154
132
106
Blasting Cap wrote:Why would Mantle die?
It is designed for use on CGN hardware....designed and made by AMD.
You think NV and AMD will hand over their latest hardware to Microsoft and then tell them to design an API to work at close to the metal level with it?
I dont think so.
I think DX will remain as the "layer above" to ensure decent compatibility across all platforms with specific low level APIs to come...first with Mantle for CGN and then an in house API for NV architecture.
Microsoft doesn't have to create a close to the metal API. They really only need to find a better multithreading solution.

It has been mentioned, that Mantle may only see about 10% gains on GPU bound games. It's biggest potential are on CPU limited games. If Microsoft can make better use of multiple threads, they'd really remove the need for Mantle. Dev's probably wouldn't bother with Mantle if all it did was give 10% boosts in performance.
 

jj109

Senior member
Dec 17, 2013
391
59
91
I remembered that my 780 supports ShadowPlay so I can record the run and not use FRAPs benchmark (it kills CPU performance :|)

So with my i7 w/o HT and locked to 2.84 (2.9 displayed)GHz, I recorded this little run:

Keep in mind that I actually had AI shooting at me.

http://www.mediafire.com/watch/55443mrvmjjzb3u

Compare to:

http://www.youtube.com/watch?v=Ms16uGxQzSY&feature=player_detailpage#t=1870

Alright boys, I think I have Mantle running on my GTX 780 :biggrin:

Do I need to underclock some more? I'm already running slower than a i5-4430.
 
Last edited:

sontin

Diamond Member
Sep 12, 2011
3,273
149
106
AMD used a Kaveri and a 290X.

Right now we will not really see a increase in the frames with a processor which doesn't imitate cpus from 2008.
 

jj109

Senior member
Dec 17, 2013
391
59
91
AMD used a Kaveri and a 290X.

Right now we will not really see a increase in the frames with a processor which doesn't imitate cpus from 2008.

I thought an A10-7850K could match the slowest i5 in multithreaded? My setup was even slower than the slowest Haswell i5.
 

sontin

Diamond Member
Sep 12, 2011
3,273
149
106
i5 has 4 cores, Kaveri only 2C/4T.
Even in multithreaded applications a i5 will be around 40%+ faster than Kaveri.
 

krumme

Diamond Member
Oct 9, 2009
5,956
1,595
136
Microsoft doesn't have to create a close to the metal API. They really only need to find a better multithreading solution.

It has been mentioned, that Mantle may only see about 10% gains on GPU bound games. It's biggest potential are on CPU limited games. If Microsoft can make better use of multiple threads, they'd really remove the need for Mantle. Dev's probably wouldn't bother with Mantle if all it did was give 10% boosts in performance.

Carfax83 link to xbox dev explains ms perspective on this and how they solve it.

Gpu rendered frame to frame tend to be more consistent than cpu.

And the lowest fps and smoothness defines experience.

So what mantle and the console adresses is vital for experience. The solution seems to be on the software side exploiting the hw.

Explained in link. Nice but long read.
 

jj109

Senior member
Dec 17, 2013
391
59
91
i5 has 4 cores, Kaveri only 2C/4T.
Even in multithreaded applications a i5 will be around 40%+ faster than Kaveri.

Ok, I loaded up some BF4 CPU benchmarks and turns out I was 10% too fast. I dropped the CPU down to 2.6 GHz no HT and ran the whole sequence:

http://www.mediafire.com/watch/28820jqn83dkdw9

The only time I dropped into the 60s was when the wave made a huge mist in the FOV, but the Mantle demo did that too, so that's not the 45%. In fact, the AMD demo didn't seem to have AI, so they weren't taking that significant penalty (see 80+ fps w/ AI vs 100+ fps after killing the enemies)

Look at the section of the run after I kill all the units: it's doing 100+ just like the AMD demo.

AMD looked like they were using a capture card too, since they had tearing in their BF4 demo. I'm still using ShadowPlay and still taking the minor performance penalty.
 
Last edited:

BallaTheFeared

Diamond Member
Nov 15, 2010
8,115
0
71
i5 has 4 cores, Kaveri only 2C/4T.
Even in multithreaded applications a i5 will be around 40%+ faster than Kaveri.

If FP mattered that much AMD design would be much slower and considerably more stupid than it already is.

Calling it 2c/4t is just sad at this point.

Threadcrapping is not allowed.

- Elfear
 
Last edited by a moderator:

jj109

Senior member
Dec 17, 2013
391
59
91
if fp mattered that much amd design would be much slower and considerably more stupid than it already is.

Calling it 2c/4t is just sad at this point.

4i/2f/4t?

Anyhow, I've made my lovely i7 as slow as a FX-4300 and I still couldn't find the 45%... Instead I found ~0% over DX11 give or take a few. Someone want to try with their 290X?
 
Last edited:

Paul98

Diamond Member
Jan 31, 2010
3,732
199
106
Ok, I loaded up some BF4 CPU benchmarks and turns out I was 10% too fast. I dropped the CPU down to 2.6 GHz no HT and ran the whole sequence:

http://www.mediafire.com/watch/28820jqn83dkdw9

The only time I dropped into the 60s was when the wave made a huge mist in the FOV, but the Mantle demo did that too, so that's not the 45%. In fact, the AMD demo didn't seem to have AI, so they weren't taking that significant penalty (see 80+ fps w/ AI vs 100+ fps after killing the enemies)

Look at the section of the run after I kill all the units: it's doing 100+ just like the AMD demo.

AMD looked like they were using a capture card too, since they had tearing in their BF4 demo. I'm still using ShadowPlay and still taking the minor performance penalty.

perfoverlay.drawgraph 1
 

3DVagabond

Lifer
Aug 10, 2009
11,951
204
106
4i/2f/4t?

Anyhow, I've made my lovely i7 as slow as a FX-4300 and I still couldn't find the 45%... Instead I found ~0% over DX11 give or take a few. Someone want to try with their 290X?

So, maybe it's not ~45% faster because of the CPU?

Although the fact that they used the Kaveri APU does add fuel to the fire that it will mostly help in CPU limited conditions. Who knows?
 

jj109

Senior member
Dec 17, 2013
391
59
91
So, maybe it's not ~45% faster because of the CPU?

Although the fact that they used the Kaveri APU does add fuel to the fire that it will mostly help in CPU limited conditions. Who knows?

That doesn't even make sense.

perfoverlay.drawgraph 1

You're lucky I still had BF4 open.

http://www.mediafire.com/watch/21wlc2s5pwzdjw2

Each horizontal tick is 3ms, so 99% of the little spikes can't be seen since they get smoothed out by the monitor. There's big section where CPU rises, but that happens in Mantle also. Starting a splash causes a 8ms hiccup. Maybe this is why [H] said the 290X was smoother than the GTX 780? Who knows.

Edit:

Hey! Looks like my i5 @ ~2.8 GHz estimation for a 7850K was right after all.

http://www.hardcoreware.net/kaveri-review-a10-7850k/4/
 
Last edited:

DamnedLife

Member
Dec 26, 2013
101
0
0
These are still all guesstimate benches. When mantle comes out there will be plenty of real benches. Mantle vs dx on i5 i7 fx with 270 280 290 290x. Just for 290X there will be 64 results for high end i7 i5 fx low end i7 i5 fx and all these for dx vs mantle. So keep your benches to yourself.

Warning issued for inflammatory language.
-- stahlhart
 
Last edited by a moderator:
Aug 11, 2008
10,451
642
126
So where is the patch? So much for it coming out with kaveri. Are they just going to drop it without any hard release date? I don't really play the game, are the dx issues resolved yet?
 

bystander36

Diamond Member
Apr 1, 2013
5,154
132
106
Carfax83 link to xbox dev explains ms perspective on this and how they solve it.

Gpu rendered frame to frame tend to be more consistent than cpu.

And the lowest fps and smoothness defines experience.

So what mantle and the console adresses is vital for experience. The solution seems to be on the software side exploiting the hw.

Explained in link. Nice but long read.
I'm not sure how anything you wrote pertains to what I wrote.

I'm just saying that on the GPU side, DX isn't that much slower. DX could still be made good enough by simply adding multithreading, and perhaps reduce some overhead on each call.

To the metal is faster, but you can see that to the metal on the GPU side of things, doesn't make much difference. It may not make all that much difference on the CPU side either if a few things were fixed.

I would like to read the link you referred to, it does sound interesting.
 

krumme

Diamond Member
Oct 9, 2009
5,956
1,595
136
I'm not sure how anything you wrote pertains to what I wrote.

I'm just saying that on the GPU side, DX isn't that much slower. DX could still be made good enough by simply adding multithreading, and perhaps reduce some overhead on each call.

To the metal is faster, but you can see that to the metal on the GPU side of things, doesn't make much difference. It may not make all that much difference on the CPU side either if a few things were fixed.

I would like to read the link you referred to, it does sound interesting.

http://www.eurogamer.net/articles/digitalfoundry-the-complete-xbox-one-interview
 

dguy6789

Diamond Member
Dec 9, 2002
8,558
3
76
To the metal is faster, but you can see that to the metal on the GPU side of things, doesn't make much difference. It may not make all that much difference on the CPU side either if a few things were fixed.

When you look at Skyrim and Bioshock Infinite and how well it runs and looks on an Xbox 360(console from 2005) vs a computer from 2005 even if it was a $3000 gaming rig(spoiler: it's not even close) you will better appreciate the difference between to the metal coding and DirectX overhead. You're very mistaken, it's an absolutely massive difference.
 

bystander36

Diamond Member
Apr 1, 2013
5,154
132
106
When you look at Skyrim and Bioshock Infinite and how well it runs and looks on an Xbox 360(console from 2005) vs a computer from 2005 even if it was a $3000 gaming rig(spoiler: it's not even close) you will better appreciate the difference between to the metal coding and DirectX overhead. You're very mistaken, it's an absolutely massive difference.
I don't own Bioshock, but Skyrim runs extremely great on the PC. At least on a decent one. However, in the early days of Skyrim, it had CPU limiting problems, the same problems I said MS needs to fix.

A crappy PC will expose its problems far earlier than a console, because mouse is far more sensitive.
 

jj109

Senior member
Dec 17, 2013
391
59
91
These are still all guesstimate benches. When mantle comes out there will be plenty of real benches. Mantle vs dx on i5 i7 fx with 270 280 290 290x. Just for 290X there will be 64 results for high end i7 i5 fx low end i7 i5 fx and all these for dx vs mantle. So keep your benches to yourself.

Almost 5000 posts of baseless speculation and you're still willing to sit on your hands? If you aren't willing to put evidence on the table when all the tools are available, don't even bother coming at me.

At least someone here has a 290/X and BF4 right?
 
Status
Not open for further replies.
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |