AMD : Battlefield 4 Patch, Mantle API and Catalyst 14.1 Update

[Paul98]

What is going to be nice is being able to upgrade only a graphics card and be able to play mantle games with good FPS.

 

[krumme]

What is going to be nice is being able to upgrade only a graphics card and be able to play mantle games with good FPS.

Cheap, fast and easy. Game is pratically the only reason i upgrade cpu/mb/ram.

 

[Despoiler]

I think unparking cores is better than disabling HT. I know with a 280x the stutter was pretty bad in crossfire. Going single card mode induced stutter as well. Rather than tink around with disabling HT, IGPs or other display devices, AMD just needs to fix their drivers and mantle implementation. It needs to be done on a quicker rotation too...

Judging by AMD's past, it's going to be a long wait until these issues get addressed.

It goes back to something I originally said before their drivers came out... AMD has a track record of not fixing problems and leaving you out to dry with promised fixes. 7xxx/280x owners continue to feel it.

If AMD wants mantle to be successful, they need to begin issuing driver fixes OFTEN. Once a month or quarter won't cut it.

I'll let all of you enjoy the growing pains while they sort this out as I'm throwing in the towel after so many "just wait" episodes.

I'm not sure where you've been looking for drivers, but AMD releases multiple drivers of the same release. 13.11 had 12 different versions that came out in 3 months.

 

[PPB]

There was a performance penalty for enabling HT in the Frostbite 2 engine, is this still prevalent in Frostbite 3?

Nope, that only shows that for a 6 thread scaling engine like frostbite 2, it was wiser to disable HT on the extreme line of chips (6c/12t ->>> 6c/6t) for better performance, but not on 4c/8t chips, those would still benefit from having HT on.

 

[Venomous]

I'm not sure where you've been looking for drivers, but AMD releases multiple drivers of the same release. 13.11 had 12 different versions that came out in 3 months.

And? Did they address frame pacing? Not a chance.

 

[Despoiler]

And? Did they address frame pacing? Not a chance.

Your premise was that AMD needs to release drivers more than once a month or quarter and I pointed out that their last beta set had 4 updates for each month for that build. That means your premise is incorrect. Also, 14.1 has phase *2* of frame pacing. 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. Like I said, not sure where you are looking for drivers or information on them, but you are not getting the current information.

 

[Red Hawk]

Your premise was that AMD needs to release drivers more than once a month or quarter and I pointed out that their last beta set had 4 updates for each month for that build. That means your premise is incorrect. Also, 14.1 has phase *2* of frame pacing. 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. Like I said, not sure where you are looking for drivers or information on them, but you are not getting the current information.

Have any benchmarks been done for the Dual Graphics framepacing with 14.1?

 

[Despoiler]

Have any benchmarks been done for the Dual Graphics framepacing with 14.1?

I haven't seen any websites that have tested frame pacing phase 2. They are so concentrated on Mantle they forgot to do a full driver run down.

 

[lyssword]

Mantle REALLY boosts those with core i3 or dual core processors
http://pclab.pl/art55953-4.html
in case of i3 4330 (3.5ghz) performance goes from 43 to 78 FPS almost matching 4570k
fx 6350 also decent jump from 39 to 70 FPS

 

[BrightCandle]

Pcper.com tested the frame pacing the day or so the driver came out. Works but like I said before still no dx9 support.

 

[Zstream]

Pcper.com tested the frame pacing the day or so the driver came out. Works but like I said before still no dx9 support.

Don't expect any either.

 

[Black Octagon]

^ what do you mean? Have you read something that suggests AMD is giving up on DX9 frame pacing?

 

[zebrax2]

This quote from here Suggest that there is a plan to release DX9 frame pacing

One thing to note with phase 2 is that like phase 1, these frame pacing improvements are solely for Direct3D 10+. Direct3D 9 and OpenGL are not covered by phase 2 under single-monitor or multi-monitor configurations, and as such still use AMD’s old frame pacing algorithms. AMD will ultimately be releasing a phase 3 driver to handle these APIs, though AMD isn’t providing a schedule for phase 3 at this time

 

[el etro]

Mantle REALLY boosts those with core i3 or dual core processors
http://pclab.pl/art55953-4.html
in case of i3 4330 (3.5ghz) performance goes from 43 to 78 FPS almost matching 4570k
fx 6350 also decent jump from 39 to 70 FPS

Gaming still relies more on per core performance. Then HT gives a good boost in handling threads.

 

[IEC]

And? Did they address frame pacing? Not a chance.

I have had Radeon 290s since shortly after launch and I have had zero, zip, nada issues with frame pacing in Crossfire. It's simply not a problem with R9 290/X.

 

[Venomous]

I have had Radeon 290s since shortly after launch and I have had zero, zip, nada issues with frame pacing in Crossfire. It's simply not a problem with R9 290/X.

Well were talking about frame pacing drivers for dx9-11 for prior cards up to the R9s that were promised and promised and promised...

At this point I would expect their newest cards to support frame pacing. Still doesn't excuse their past broken promises now does it? Glad your new cards have no problems in crossfire...

 

[BrightCandle]

Back in January 2013 AMD said it would be a few months before all the frame pacing issues were fixed for their 7000 series cards. Its been a year and they have fixed most of it now and I believe its just DX9 that is outstanding. Its still not quite as good as Nvidia's frame pacing most of the time but by most accounts its dramatically better than what it was before it was released. With just DX9 remaining that is an improvement but I suspect the Skyrim fans are still annoyed its not been done.

As I understand it Mantle doesn't have frame pacing yet and its one of the reasons Mantle doesn't support crossfire yet. Anyone heard of a date/timescale of when its coming and for which cards?

 

[Elfear]

Back in January 2013 AMD said it would be a few months before all the frame pacing issues were fixed for their 7000 series cards. Its been a year and they have fixed most of it now and I believe its just DX9 that is outstanding. Its still not quite as good as Nvidia's frame pacing most of the time but by most accounts its dramatically better than what it was before it was released. With just DX9 remaining that is an improvement but I suspect the Skyrim fans are still annoyed its not been done.

Indeed.

I haven't heard any firm date on phase 3. Hopefully soon.

 

[Red Hawk]

Well were talking about frame pacing drivers for dx9-11 for prior cards up to the R9s that were promised and promised and promised...

At this point I would expect their newest cards to support frame pacing. Still doesn't excuse their past broken promises now does it? Glad your new cards have no problems in crossfire...

Enhanced AMD CrossFire™ technology frame pacing – Support for 4K panel and AMD Eyefinity and AMD CrossFire™ solutions (including the AMD Radeon™ R9 280, 270 Series, HD 7900 Series, HD 7800 Series) and Dual Graphics configurations​

Cards prior to the 290 series do have framepacing. Heck, the initial framepacing driver came out before the 290. Unless you're talking about pre-GCN graphics cards, in which case IIRC pre-GCN cards never really had the same framepacing issues. The issues were mainly single-card GCN stuttering (resolved with an update early last spring), crossfire GCN stuttering (addressed with an update in the summer and continuing to be addressed here), and stuttering on Dual Graphics setups with Trinity (presumably addressed here).