OFFICIAL: BF4 Mantle Cat 14.1 Performance and Feedback Thread

Twizty said:
Thought I'd throw in my 2 cents worth since I'm trying to figure out how I'm seeing people with similar systems post gains with Mantle, and yet I'm seeing a loss.

PC Stats:
Core i7 2700k stock speed
16gb Kingston PC3-10700
Asus Maximus IV Extreme-Z
Sapphire HD 7770 1GB stock speeds

The data: (apologies for the largeness of the images)

BF4 set to Medium preset

Average FPS
Avg.FPS.Chart.PNG


FPS Linechart
FPS.Linechart.PNG


Frame Times Linechart
Frame.Times.Linechart.PNG


And if anyone wants to see all the data collected, here's my spreadsheet.
https://dl.dropboxusercontent.com/u/37751736/Mantle/Zavod.311-Comparison.Data.xlsx

Any input is welcome.
Click to expand...

what resolution are you running? I'm gonna guess 1600x900?

Your 7770 only has 1GB of VRAM. With the fact that Mantle seems to use more VRAM for the same settings, it's probably pegging your VRAM really hard, dropping performance really badly. It also may just blatantly have zero optmization for the 7770/7790. Who knows.

I'd monitor your VRAM in the test server with GPU-Z. maybe drop to 720p to see if you can free up some VRAM, if it indeed is flooding. see how that goes.
 
EnhancedInterrogator said:
According to what I've read, the only three cards that are fully supported (right now) are the 260, 290, and 290x, in a single card configuration. I think anyone else reporting gains hasn't really tested their card in a large multiplayer match, or they run at really low settings that hardly use any VRAM.
Click to expand...

I beg to differ.

Using these settings on a 7870 overclocked to 1150 core 1325 memory:
ScreenshotWin32-0002.png


I gained over 25fps on my average framerate, during 64 player siege of Shanghai, tested on the same server. Even with all the hitching, which was particularly frequent and occasionally would drop several frames, Mantle still turned in a much higher average. I run a Phenom II X6 which I didn't think was holding me back all that much. It seems my 7870 still have plenty to give!

It still stutters a bit in test server, even though the overall average is barely higher. The CPU demand is way lower in the test server, so it makes sense. I imagine the VRAM reqs are lower, as well. In the test server, Mantle basically pegs my 2GB VRAM capacity. I haven't tried to track VRAM in multiplayer yet.

This is what my test server lap looks like. You can stream it or download the source file, which looks a bit better than the stream. It's meant to be a quick and dirty encode, however.
https://www.dropbox.com/s/wgd3gr8fdlmn429/bf4testserverrunsamplea.mp4

chameleoneel said:
I finally bothered to calculate my average frame rates


Earlier I posted raw perforoverlay frame data (CSV files) for 13.12whql, 14.1 DX11 and 14.1 Mantle. You can see my system specs and BF4 graphics settings, and all of the other data I have compiled. Such as screenshot comparisons and VRAM usage:
http://hardforum.com/showpost.php?p=1040594668&postcount=29

There are two sets of data. One set is for a test server lap which was kept as similar as possible for each (I linked a video of what such a lap looks like).

The other set of data is for a roughly 6 minute play session, in 64 player Siege of Shanghai. (timed for no less than 6 minutes). I used a 24/7 Siege of Shanghai server so that I could ensure reliable testing of the same map on the same server. These multiplayer sessions also had AMD Performance Monitor dumping CPU core/s and GPU core utilization data. So the CPU load is slightly higher than what it normally would be.

Here are the average frame rates for the test server lap:

14.1 Mantle: 108
14.1 DX11: 103
13.12 DX11: 104

You can see here, the Test server probably doesn't stress CPU performance. As such, the average performance is basically the same. There are points at which Mantle performs better in the Test Server. But overall it is pretty much the same.

-----------------------------------------------------------

Average frame rates for 64 player Siege of Shanghai, during 6 minutes of play:

14.1 Mantle: 85
14.1 DX11: 58
13.12 DX11: 54


Here is where we get HUGE gains. That speaks for itself. Even despite the fact that the Mantle run was hitching every few seconds, with occasional hard stops/several frame drops, it still turns in a much higher average number.

The hitching is likely related to the fact that Mantle uses a lot more VRAM, for the same settings, on my system. which I show in this post here:
http://hardforum.com/showpost.php?p=1040596739&postcount=76

I have a 7870, so it is probably a bug with how Mantle is working on those cards/my system. However, it would be interesting to hear from other 7870 users about their experience, in specific regard to smoothness of play/hitching/stuttering. Or even 290x users that are running smooth, if they could show us VRAM comparisons.
Click to expand...
 
chameleoneel said:
what resolution are you running? I'm gonna guess 1600x900?

Your 7770 only has 1GB of VRAM. With the fact that Mantle seems to use more VRAM for the same settings, it's probably pegging your VRAM really hard, dropping performance really badly. It also may just blatantly have zero optmization for the 7770/7790. Who knows.

I'd monitor your VRAM in the test server with GPU-Z. maybe drop to 720p to see if you can free up some VRAM, if it indeed is flooding. see how that goes.
Click to expand...

I run at 1080p, should've mentioned that previously. I'll have to do that next time I pop in game, see what it's doing.
 
I guess "gains" was a poor word choice. I'm seeing better fps too with my 7850 under specific scenarios. But the hitching and worse minimums make the overall experience bad.

You're also using a 6-core AMD processor which is supposed to work really well with mantle. Have you tried comparing framerates when you have Ultra textures with 2x MSAA? I bet the framerate hit would be much more significant with Mantle than it would be under D3D.

Not saying you shouldn't stick with Mantle, just that it's not performing near where it supposed to once AMD finishes it. I think at that time you shouldn't have problems increasing a lot of your settings like textures and lighting or AA.
 
AND there is nothing guaranteed by AMD unless you use AMD products, support for Mantle with Nvidia or Intel stuff will come from them if it ever does, so if you see performance gains, awesome, if not, obviously you have not read the fine print "do not expect boost from anything but AMD processors, Radeon graphics cards or AMD APU products"
 
EnhancedInterrogator said:
I guess "gains" was a poor word choice. I'm seeing better fps too with my 7850 under specific scenarios. But the hitching and worse minimums make the overall experience bad.

You're also using a 6-core AMD processor which is supposed to work really well with mantle. Have you tried comparing framerates when you have Ultra textures with 2x MSAA? I bet the framerate hit would be much more significant with Mantle than it would be under D3D.

Not saying you shouldn't stick with Mantle, just that it's not performing near where it supposed to once AMD finishes it. I think at that time you shouldn't have problems increasing a lot of your settings like textures and lighting or AA.
Click to expand...

It's ok. I took issue with the "specific scenarios" comment (you aren't the only one saying essentially that) because those are my daily play settings. I didn't pic those settings special for the tests. That's what I've been playing BF4 with for the past two months.

Due to the VRAM issue, I bet Ultra and some AA would perform worse, not even as good. I'll test it. I've betting why that 7770 took a dive in those graphs a couple posts above this. Once the VRAM issue is fixed, things should be all up from there.
 
Buyers said:
Maybe its an issue that manifests itself over time? Others are speculating that its a PunkBuster issue, where its trying to take a system screenshot. Some also suspecting GPU clocks changing due to power requirements or heat.
Click to expand...

Make sure to not alter the settings or tab out of the game, after u get into the game with Mantle! there has been reports, that tabbing out of the game or alter the settings is the reason for spikes/stutters.
 
dragonstongue said:
AND there is nothing guaranteed by AMD unless you use AMD products, support for Mantle with Nvidia or Intel stuff will come from them if it ever does, so if you see performance gains, awesome, if not, obviously you have not read the fine print "do not expect boost from anything but AMD processors, Radeon graphics cards or AMD APU products"
Click to expand...

I don't think they ever pitched this to improve anything but their own products.
 
psyside said:
Make sure to not alter the settings or tab out of the game, after u get into the game with Mantle! there has been reports, that tabbing out of the game or alter the settings is the reason for spikes/stutters.
Click to expand...

The only setting i changed for testing was back and forth between D3D and Mantle between tests. What i did was have GPUZ pulled up and start its logging when BF4 starts to load. Then while in BF4 i set the in-game frame log file to run and play without interruption or alt-tabbing. I switched back and forth between Recon class and Engineer class to have a mix of long range views and up close explosions. After 15min or so or the end of a round, id stop the in-game logger and exit the game, then stop the GPUz logger. To switch API's, i'd start up the test range and change to/from Mantle there then exit, so that the API swap doesn't affect anything while actually playing on a full server during testing.
 
I ended up uninstalling the 14.1, wiping verything with driver fusion and display driver uninstaller, and went back to 13.12whql.

the 14.1 drivers are too buggy in windows, for my daily use. I was getting lots of internet video issues and sound sync issues. Today I had no sound for internet video, but sound via Media Player Classic.

I also had some display corruption in PSCX2.
 
Anyone have any info yet on the issue with BF4 crashing on the Level Loading part? I have tried every which way to install the drivers. As far as I can tell I don't have the integrated graphics enabled. I actually never even installed them to begin with.
 
If anyone is getting the memory leak bug crashing BF4 i "fixed it" buy setting system managed page file.
 
you should not have to set system managed page file, there is a reason why some of us choose to control our own, and to be fair if you set a proper sized page file ~3x installed memory, and it chews through it there is some kind of crazy leak.

As for others reporting "bugs" are you all submitting them the best you can to AMD so hopefully they can see more results on various systems/configs etc as maybe just maybe something simple like "it crashed when using this in this way" helps them get the stability up where it needs to be.

I wont bother using it right now, as I still mine on my machine and apparently performance tanks with these drivers trying to do this :(
 
chameleoneel said:
I ended up uninstalling the 14.1, wiping verything with driver fusion and display driver uninstaller, and went back to 13.12whql.

the 14.1 drivers are too buggy in windows, for my daily use. I was getting lots of internet video issues and sound sync issues. Today I had no sound for internet video, but sound via Media Player Classic.

I also had some display corruption in PSCX2.
Click to expand...
agreed, they need to fix them. I did the same because 14.1 were unstable on 3 systems
 
dragonstongue said:
As for others reporting "bugs" are you all submitting them the best you can to AMD so hopefully they can see more results on various systems/configs etc as maybe just maybe something simple like "it crashed when using this in this way" helps them get the stability up where it needs to be.
Click to expand...

For BF4 specifically, I filled out an official 14.1 support form and linked them to my master post that contains all of the data that I personally collected for BF4:
http://hardforum.com/showpost.php?p=1040594668&postcount=29

I also briefly explained to them right on the form, the kind of stuff they would find in that post. So hopefully they didn't ignore it. I have a lot of raw data there.

For general driver issues in windows, I have not yet filled out an official form. But I did post about it in the 14.1 issues/support thread here. Which I assume is being monitored by AMD.
 
Finally got around to testing the mantle driver. My config is in my sig.

@2560x1440 I saw at least a 10% performance benefit in fps, most of the time closer to 20%. With Mantle enabled the motion felt much smoother, more buttery. The FPS numbers definitely don't tell the whole story, I think the frame time improvement is noticable in large online MP sessions.

The only problem I have is hitching, where the FPS would lock up almost entirely. This was pretty rare but I had to go back to Direct 3d in BF4 because it resulted in my death on more than one occasion. I would say these FPS locks happened every couple minutes. I'm certain it's just a driver issue and I honestly can't wait for them to sort it out. With Mantle enabled I never dropped below 80 fps and usually had 110fps +. My refresh rate at 2560x1440 is 108hz so I'm excited and looking forward to the future. Best $400 graphics card I've ever bought!

I upgraded from a GTX 670. FPS almost doubled.
 
I did some testing myself on some spare parts I had laying around.

i5-750
4GB DDR3 1333
Radeon 7850 1GB




That's right, in all but one situation (Staw Swarm @ low) DX11 was outperforming Mantle. Looks like the increased memory usage of Mantle really hurts on cards with only 1GB of vram.
 
7800 series is not a good example to base off of. Mantel appears to be optimized mainly for GCN 1.1 at this time which includes R9 290/290X, R7 260/260X, Kaveri, etc. and not GCN 1.0 R9 270/270X, 7800 series, etc.

Mantle performance for the AMD Radeon™ HD 7000/HD 8000 Series GPUs and AMD Radeon™ R9 280X and R9 270X GPUs will be optimized for BattleField 4™ in future AMD
Click to expand...
 
Playing @4k with two 290Xs with the Mantle driver in BF4 resulted in crashes due to DirectX errors constantly, Sometimes before the game fully loaded, but usually after about a minute or two. Had to roll back to the WHQL driver because the last betas break Eyefinity with MST enabled.

It feels like I'm back on team green when 4k monitors first appeared 4-5 months ago. I'm not impressed so far. I hope newer betas are better.
 
Sufu said:
Playing @4k with two 290Xs with the Mantle driver in BF4 resulted in crashes due to DirectX errors constantly, Sometimes before the game fully loaded, but usually after about a minute or two. Had to roll back to the WHQL driver because the last betas break Eyefinity with MST enabled.

It feels like I'm back on team green when 4k monitors first appeared 4-5 months ago. I'm not impressed so far. I hope newer betas are better.
Click to expand...

Thanks for posting. I'm planning on switching to the Vizio 4ks for PC gaming so I'm trying to track user experience.
 
mi7chy said:
7800 series is not a good example to base off of. Mantel appears to be optimized mainly for GCN 1.1 at this time which includes R9 290/290X, R7 260/260X, Kaveri, etc. and not GCN 1.0 R9 270/270X, 7800 series, etc.
Click to expand...

I'm betting the 7850/7770 issues are the lack of ram. Mantle is bugged right now on VRAM usage. My 7870 gained over 25fps on it's average framerate, using my daily play settings for BF4 @ 1080p.
 
Does anyone know if this increased usage of VRAM is part of the design of mantle or is it a bug that will be fixed at some point? Like, is it possible that it needs more VRAM in order to use less of the CPU?
 
I asked Johan, a large vram usage increase is apparently abnormal.
 
Isn't GPU memory management in the hands of the developers now?

If so, it's likely to be a DICE problem rather than a Mantle problem. Maybe they optimized for 2GB+ vram, or maybe it really is just a bug.
 
GCN 1.0 Radeon 7k/8k GCN 1.1 7790 GCN 2.0 Rx series.
http://www.pcper.com/reviews/General-Tech/Core-Mantle-Questions-Interview-AMDs-Guennadi-Riguer
"[Guennadi] The current Mantle solution relies on the same shader generation path games the DirectX uses and includes an open-source component for translating DirectX shaders to Mantle accepted intermediate language (IL)."

So just in this alone, it could be why the amount of Vram being used has seemed to balloon, the 7870 as an example with 2gb Vram (7850 with 1, and for sure 7770 with 1) is not limited by it really cause if you add super high res textures, use lots of AA/AF it runs out of power before it really need the extra memory, 7870XT can kinda use the extra memory/bus as it does have the extra shaders and such.

Just cause a card can use very high levels of AA and "run" ultra does not mean it should be lol.
 
dragonstongue said:
GCN 1.0 Radeon 7k/8k GCN 1.1 7790 GCN 2.0 Rx series.
http://www.pcper.com/reviews/General-Tech/Core-Mantle-Questions-Interview-AMDs-Guennadi-Riguer
"[Guennadi] The current Mantle solution relies on the same shader generation path games the DirectX uses and includes an open-source component for translating DirectX shaders to Mantle accepted intermediate language (IL)."

So just in this alone, it could be why the amount of Vram being used has seemed to balloon, the 7870 as an example with 2gb Vram (7850 with 1, and for sure 7770 with 1) is not limited by it really cause if you add super high res textures, use lots of AA/AF it runs out of power before it really need the extra memory, 7870XT can kinda use the extra memory/bus as it does have the extra shaders and such.

Just cause a card can use very high levels of AA and "run" ultra does not mean it should be lol.
Click to expand...

Using my benchmarks for example, you don't need to be running high levels of AA/AF to hit the memory limit, on what would otherwise be playable in DX11. Also, considering I saw similar results in Star Swarm at any setting other than low, it doesn't appear to be exclusive to BF4 optimizations.

chameleoneel said:
My system is the following:

HD 7870 overclocked to 1150 core and 1325 memory
Phenom II X6 overclocked to 3.6ghz
8gigs of ram
Soundblaster X-Fi

Windows 7 64-bit, running BF4 in 64-bit mode.


**UPDATE

looks like the stuttering is very likely due to Mantle using quite a bit more VRAM space, for the same settings:



CONFIRMED to be an issue (at least for some of us...)

using the same system and the exact settings I I have been using all along:


here is vram usage compared between DX11 and Mantle, on my 7870:

*these shots were taken soon after loading into the test server.

DX11 screen 1
https://www.dropbox.com/s/352up4h89lvcq90/dx11vmemtest1.jpg

Mantle screen 1
https://www.dropbox.com/s/bmarkleomxbinkp/Mantlevmemtest1.jpg

DX11 screen 2
https://www.dropbox.com/s/oo2xpxwo1ed6von/DX11vmemtest2.jpg

Mantle screen 2
https://www.dropbox.com/s/3u3h6e7oxkm309b/Mantlevmemtest2.jpg

*also captured in these screenshots, Mantle makes my card run way hotter.

**I took a few more screens. This time after I'd been in the server for several minutes. I loaded in, ran my usual lap (which takes a few minutes) and then took a couple of shots. Vram usage is even higher. DX11 temps top out at 58. Mantle pushes it even higher still...
Click to expand...

According to these result, Mantle is currently using roughly twice the vram of DX11. I really hope this is a major bug they can sort out, and not a sign of things to come. Alternatively, this may be a justification for all those low-mid range cards with double the normal amount of vram ;)
 
I think part of the "issue" if you will is the way Mantle is right now for example in BF4, in which the calls are still made Via DX but patched through to a mantle layer (so like an intermediate layer so the devs do not have to do anything fancy to get some of the performance that mantle has and not need to take the time to learn a new language) so very likely it is using more memory as it "stacks" the amount needed or some such thing, so in this at least what I have been reading as the devs and possibly AMD directly can get more familiar/comfortable with the way it needs to be done it should very well bolster performance by a very large margin, for a first showing I daresay quite amazing really.

1.5Gb on a 2gb card to my mind is fine, it means it is using more available memory BUT if the performance is stutter or not smooth then obviously this is an issue, and lets face it DICE might be great at trying new things out, but bar none are horrid at making it work well.
 
dragonstongue said:
I think part of the "issue" if you will is the way Mantle is right now for example in BF4, in which the calls are still made Via DX but patched through to a mantle layer (so like an intermediate layer so the devs do not have to do anything fancy to get some of the performance that mantle has and not need to take the time to learn a new language) so very likely it is using more memory as it "stacks" the amount needed or some such thing, so in this at least what I have been reading as the devs and possibly AMD directly can get more familiar/comfortable with the way it needs to be done it should very well bolster performance by a very large margin, for a first showing I daresay quite amazing really.

1.5Gb on a 2gb card to my mind is fine, it means it is using more available memory BUT if the performance is stutter or not smooth then obviously this is an issue, and lets face it DICE might be great at trying new things out, but bar none are horrid at making it work well.
Click to expand...

No

This would not work because what you are describing is called a wrapper and wrappers do not make games faster. There is no new language to learn.

The developers are now completely in control of what there doing on the video card for the most (if not all) part. So problems with ram usage is from DICE code.
 
http://www.pcper.com/reviews/General-Tech/Core-Mantle-Questions-Interview-AMDs-Guennadi-Riguer did you not read this?

Anyways does not matter, as it stands, Mantle and BF4 as a very basic example, takes the same path they are used to, but applies it differently which they can then optimize the way they see fit.

So it seems to be in this specific case not a 100% new language to learn, or I bet you anything the visuals would not be the way they are. Before you are snide about that, DICE/EA are all about the fancy glitzy look these last couple of years, they certainly would not tone it back on purpose especially if it was built from the ground up as they keep hinting at as it has been a "work in progress for quite awhile"

Maybe when it comes to a Linux implementation or a game built from the ground up using mantle it might be different.

Did Creative Alchemy not allow the hardware to be used in a much different fashion then the software allowed due to limitations? Yes it is very much a wrapper as it takes 1 call and makes it to another.

So Mantle CURRENTLY could very much be exactly that and the article linked seems to be hinting at that, instead of doing A I want you to do B via Mantle instead of DX does not sound like a complete overhauled language to me, sounds like a code branch.
 
hes talking about compiling the shaders during build time, not interpreting in real time.

lol
 
dragonstongue said:
GCN 1.0 Radeon 7k/8k GCN 1.1 7790 GCN 2.0 Rx series.
http://www.pcper.com/reviews/General-Tech/Core-Mantle-Questions-Interview-AMDs-Guennadi-Riguer
"[Guennadi] The current Mantle solution relies on the same shader generation path games the DirectX uses and includes an open-source component for translating DirectX shaders to Mantle accepted intermediate language (IL)."

So just in this alone, it could be why the amount of Vram being used has seemed to balloon, the 7870 as an example with 2gb Vram (7850 with 1, and for sure 7770 with 1) is not limited by it really cause if you add super high res textures, use lots of AA/AF it runs out of power before it really need the extra memory,
Click to expand...

dragonstongue said:
1.5Gb on a 2gb card to my mind is fine, it means it is using more available memory BUT if the performance is stutter or not smooth then obviously this is an issue, and lets face it DICE might be great at trying new things out, but bar none are horrid at making it work well.
Click to expand...

Ok so you are disregarding some realities about GPU peformance and also disregarding some things that have been shown in this thread, with actual data. I'm sorry if that sounded mean, I'm not trying to be.

First of all, it's true that if you push resolution really high, or use a lot of AA, a videocard can run out of performance, before it runs out of VRAM.

But I don't think anyone in this thread has presented data, in which the videocard was maxed out on performance in DX11 and then complained about the Mantle experience with those same super high settings.

With my own test for example, I used my everyday play settings. Which are well within the performance limits of my GPU. I can actually play 1080p Ultra just fine, on my 7870, with good framerates. But to ensure a good minimum framerate even when things get crazy on screen, I use mixed settings, still 1080p.

ScreenshotWin32-0002.png


In the test server, those settings under DirectX give me about 105fps average, with about 1100mb of max VRAM usage.

With Mantle, the framerate is about the same (because the test server isn't CPU limited), but it uses over 1700mb of VRAM. SAME SETTINGS. and it hitches and stutters.

Those same settings, with DirectX, get me about 58fps average; during 64 player Siege of Shanghai.

With Mantle, I actually got a lot more performance in regards to framerate (which means my GPU was not maxed out to begin with. In fact, it had a lot more to give, but needed more from my CPU), 85 fps average, on 64 player Siege of Shanghai. But the stuttering was really bad. Every 3 seconds it would stutter and hitch. Occasionally, I would get a hard hitch, where the game would stop and skip a few frames, then I would move again. (I did not test VRAM usage in Siege of Shanghai. But I can assume it was really bad, due to awful hitching and due to the fact that even the test server was nearly maxing out my VRAM in Mantle, with optimized settings. Not even close to pure "Ultra")

*you can see all of my raw and detailed data here, as well as other info:
http://hardforum.com/showpost.php?p=1040594668&postcount=29

So with my tests, it's not the case that the GPU is already being worked hard, as you talked about. and as I said, the other data in this thread is usually showing GPUs that aren't being overworked to begin with. and their VRAM usage is well within limits. But high VRAM usage doesn't necessarily mean the GPU is being maxed out on peformance. A moderate resolution, no AA, but really large textures can use a lot of VRAM. Texturing is easy for modern video cards. It really doesn't matter how high resolution they are, as long as they fit into VRAM.

But then you look at the Mantle data and things go wrong. Setups where the GPU performance is not being maxed out and the VRAM usage is well under the limits of the card in DirectX. But those same settings in Mantle; people are showing higher framerates, but reporting stuttering and hitching. Because Mantle currently increases the VRAM usage to max or near max, for their card. So even though the GPU core is performing well, the ram usage is too high.
If you look closely at my VRAM data, you see two numbers: 1. is overall VRAM usage "dedicated". This is probably the all of the main, permanent assets for the level/map/etc.

2. is "dynamic" VRAM. This is probably what the Videocard is actually processing at any given moment. As you move around, shaders are being processed, AA, Tesselation, LoD changes, etc.

With my tests, Mantle uses about 1700mb of dedicated ram. But at any given moment, it is changing/swapping between 200 and 500mb of VRAM. That often puts it over the limit of 2000mb for my card and is the reason for the stuttering and hard hitching. It's not way over the limit, however. So overall performance is still higher. Likewise, DX11 had dynamic ram usage between 100 and 300mb.


Someone else posted performance numbers for their 7770 1GB. The DX11 numbers were good numbers, showing that the settings were not maxing out the GPU.

The Mantle numbers showed a HUGE drop in overall performance. I suspect that the VRAM increase was too significant. The card was probably swapping VRAM data constantly, causing overall performance to stay lower. Rather than just occasionally stuttering under VRAM spikes.
 
Question then, dynamic is memory that "moves around" or is able to be used at will sort of speak where dedicated is just that already in use/called for, is it not something like windows in that it takes away from total memory and is giving X to dedicated and X to dynamic so in this case if you have 2gb Vram you could not exceed this amount?

I know that the memory bus can only pass so much at one time and the memory might be very fast but is not passing a lot of this raw speed through the bus to the display, so maybe it is having issues by having not enough dynamic or trying to feed too much through the already constrained bus?

Just wondering I suppose.
 
Any word yet on a fix to the crash when trying to load a level? I still haven't been able to load into a single map with mantle.
 
Use system managed page file, it helps to avoid errors, for me it resolved all of the issues.
 
Patch is out today. Seems there was a memory leak:

Mantle
-Fix for a crash that would occur when activating full screen in portrait mode
-Fix for stuttering that could appear during video sequences on multi-GPU PCs
-Fix for a memory system leak that could cause stalls, which would result in frames taking longer to process
-Reduced the amount of stalls that occurred when running with high graphics setting that require more GPU memory than is currently available
-Fixed screenshots on multi-GPU PCs
 
Feb 13 PC Game Update Notes
-General stability improvements
-Fix for an issue where spawning into, or switching to, a gunner seat in an IFV/MBT sometimes could cause the game to crash
-Fix for missing sound in Team/Squad Deathmatch
-Fix for an issue in the Defuse game mode, where a bomb carrier would be permanently spotted
-Decreased the rate at which the kill card would incorrectly display 0 health, while the enemy was clearly alive
-Fix for an invisible wall that was incorrectly present in one of the fallen concrete pipes on Zavod 311
-Fix for an issue where bullet impact sounds weren’t properly matching the actual number of impacts
-Fix for an issue where the “Draw” message would not display on-screen once a Conquest round ended with both teams having the same amount of tickets
-Fix for an issue where long IDs wouldn’t scroll on dog tags
-Fix for missing grass physics in terrain
 
I don't play BF4 much, how can I check that it's updated? I just loaded up origin and right clicked on bf4 and did "check for update" but it says its already up to date. I doubt that unless the patch was really small or something today.
 
fixes don't have to be really really big ;) unless you are called MS then patches will be very big ;).
It should be active .
Just did some TDM and it wasn't as bad (fixed?) sometimes with the framedrop as it used to be.
 
You must log in or register to reply here.
Back
Top