"The numbers suggest the Mantle build delivers bigger gains in CPU-limited scenarios. According to EA DICE, the 58% performance increase on the CrossFire-equipped system is due to the CPU being too slow to feed the dual 290X cards. This may be the first time I've seen a six-core Sandy Bridge-E processor referred to as too slow for anything." This low level Mantle API behave exactly the same, as API used in consoles which have relatively starved CPUs. I believe, that rfactor2 will be not getting much from Mantle(mayority), as most modern cpu's buyed today is up to task and at top of that, even latest AMD gpus has problem to running rfactor2 smoothly. The irony in which i believe is that Nvidia cards probably will be giving better result if Nvidia will decide using and participated with Mantle API. So quite opposite to situacion for what was Mantle made for. On the other hand, i was try AC today and with 24 cars i(and most of others) get warning of CPU overload(95%). Despite that this can be happen because of lack of optimization, i believe AC will quite possibly have more advantage with Mantle as rF2.
Not sure if you understand what Mantle is, rF2 won't get anything from Mantle unless ISI support it (don't hold your breath). I don't have AC (yet) but interestingly you're the 2nd person today who's indicated that AC seems more CPU bound than rF2 so yes AC could have more to gain but only *if* Kunos supports it.
Definitely i don't understand technicalities of it, but i sure know that also game developer must support it. I just don't mention it because is obvious fact.
Not really. If you develop for Windows only, DirectX is installed on every machine. Mantle is a DX alternative. If you use Mantle, you DON'T use DX. If we added it, it would mean another pipe and another option for us to deal with, and users to deal with. Not that this is always a bad thing.
Some misunderstanding here Maybe my writing Game developers must support it, IF they want users to have advantage from Mantle.
Yep. It's not any reason to switch manufacturer or think any game not using Mantle will work any better. Obviously it would be very nice if every dev started to support it though. Better for the industry, certainly... When there is a real 'war' between AMD and nVidia not only in hardware, but drivers and APIs, that's going to move the industry forward much more.
Good post -> http://isiforums.net/f/showthread.php/18087-Experimenting-with-AMD-graphics-card-settings/page2#24
Ok, driver is posted up, and I have an independent review from guru3d. Reminder, this is all beta, and in release notes, it states they will optimize even more in future driver releases. Download: http://support.amd.com/en-us/download/desktop?os=Windows 8 Release Notes: http://support.amd.com/en-us/kb-articles/Pages/latest-catalyst-windows-beta.aspx Mantle FAQ: http://support.amd.com/en-us/kb-articles/Pages/mantle-faq.aspx review: http://www.guru3d.com/articles_pages/amd_mantle_preview,2.html Even at 1440p, it gains 10% on DX11. PC is, i7 3960x, R9 290x, 2560x1440, ultra quality, and 4xAA
there is a demo (star Swarm stress test) in steam by oxide games, mantle and dx11 suport, the difference is brutal.
The renderer reduce CPU/API bottleneck by a huge margin, won't that help simulation game like rFactor 2 where i kept getting 70%-80% GPU usage because of CPU/API bottleneck?