Mantle Feedback/Bugs

[LtMatt]

To be fair just because the thread name changed shouldn't mean that topics of discussion should be altered/stopped. The mods did specifically direct people to this thread when they locked the other one for being too long.

That's fine but we have a new Mantle benchmark thread here. That thread was created because the original benchmark thread (this thread) got derailed and ruined. The new thread was created with the approval of the dons. So i turned this one into a Mantle feedback and bug thread.

 

[LtMatt]

lol indeed . But irrespective of comparisons and what not it still should be (in the context of Mantle) that if you're graph dropping in a thread where the graph is not explanatory as to its origin then a one line caveat is both polite and necessary. As said before, you know as well as I do that when graph dropping hardly anyone clicks the actual article link unless they're looking for holes to pick at.

Herein lies the problem. If people had actually looked at the article, read the link i gave them and watched the benchmarking sequence they used, then we likely wouldn't be having this debate.

Thing is matt, this isnt your forum, people are free to raise whatever points they want, as long as they dont start throwing needless insults in

Raising a point that a posted set of info might be unrepresentative of general gameplay was a totally valid point, considering it was not mentioned within the post that originated the info

You're having a pop at rusty for not letting itgo, but it takes two to tango

I never said it was my forum. I was just sticking to the topic, posting a benchmark with a link to an article about Mantle and its performance vs DX.

It was not mentioned because i linked to the article, which contained the link to the video. It was a 30 second sequence of a run through the resource heavy part of the map, on foot. A good indicator of performance as its pure gameplay and not sitting on the spawn screen or waiting in base that would have been part of pgi's benchmark comparison. Again this comes back to the reason of why it was not a particularly relevant comparison to what they were testing. Its basic benchmarking procedure.

I only responded to Rusty telling me to stop being rude. I just replied explaining why it was not a relevant comparison and that was not actually being rude just stating facts based on the different benchmark comparisons.

 

[LtMatt]

Re-read the post you've quoted.

You're hiding behind the fact you linked it but as I've already said if you're re-posting / graph dropping on another forum (here) then it's your responsibility to represent the data in a way in which it was intended to be viewed. You didn't do this hence the issue. A 1 line caveat would have sufficed. Oh well, move on. Perhaps just bear that in mind for the future.

No there was no need for a caveat. If the person is interested in the graph its up to them to read the article and scrutinise it if needs be. I'm not going to copy and paste the whole thing for them to read, that's up to the person who feels the need to challenge the data, for whatever reason. I provided the source in the original post so it was not just a random graph without a link. Remember Rusty in future if you want to pick holes in something it pays to read the article first before casting judgement.

 

[LtMatt]

I feel for you Matt i really do, these threads are just destined to got *** ups.

Its not as if i was even bad mouthing Nvidia performance. It was a benchmark about a 290X and Mantle Vs DX. Groan.

 

[LtMatt]

If someone feels the need to challenge the data for whatever reason its up to them to read the article in question. I provide the graph and the source, like i always do. Nothing is going to change on that front.

I watched their benchmarking sequence and i feel it is representative real world performance you're likely to get on that map in that area. I'd rather see a intense, repeatable, short and consistent benchmark than a long drawn out bench that is going to be inconsistent and hard to reproduce reliably. When trying to highlight the performance of Mantle we want consistency and in this instance this benchmark definitely provided it. They even uploaded it so we can copy it. This is great so we can copy it if needs be and share results. Its also why it was a pointless comparison comparing it to pgi's results. I mean no harm to pgi when i say that so i hope hes not still offended by it. Nothing wrong with his benchmark at all, it just can't be compared reliably to this one.

 

[LtMatt]

Well that is a shame as it means that in certain cases there will continue to be misunderstandings for the reasons mentioned before. It's not difficult to post a 1 line caveat where required (like this case).

I think it's quite useful actually but that's your call.

Yes if someone wants to scrutinise something or find fault with it, the responsibility lies with them to do so by reading the linked article.

What's your thoughts on Mantle then Andy after using it.

Interested to hear both Andy's and Rustys feedback on Mantle.

 

[LtMatt]

'Find fault' is not actually what happened though is it? Don't be so defensive over a benchmark - it's not yours!

I have no opinion having not used it but looking at the data it's difficult to see how much benefit it's going to have for your average high end GPU user with an overclocked i5/i7 who uses 1080/1440 on max settings.

Find fault / scrutinise / not representative. All the same. Its not my benchmark, but im just explaining the context of it and why its not reliable to compare it to something else.

If you have no opinion having not used it then >> time to leave the thread. Perhaps the Mantle Benchmark thread would be better for you. If you actually check that out, you'll see real world examples of the performance difference in exactly the scenario's you mentioned, regarding the components you mentioned. I saw a 14% increase with an i7 and a single 290 card in the Mantle benchmark thread over the same benchmarked scene on DX.

 

[LtMatt]

You posted a benchmark here and it was commented upon here. If you didn't want that to happen here then you should have posted it in the other thread. Simples .

My apologies if my posts were not under the umbrella of 'feedback'.

No problem with talking about the benchmark, it was posted here to be discussed. Its also something us Mantle users can try and copy and replicate and see what results we get. However you spoke about wanting to see real world results, i just informed you that you've been looking in the wrong place for them and derailing the thread somewhat in the process.

 

[LtMatt]

Nicely done the CPU spike's are a real pain, I have no idea if this is Game or Driver. But what I do know is I cant wait for 14.2 or next Patch.

Until the stutter and crashing is sorted I dont think we can give a true test on Performance. For example the CPU spikes make the frame rate drop thus giving false reading.

+1.

It gave us a nice taster, but we want moar. Need to get crossfire working properly as well.

 

[LtMatt]

So i'm back after my little holiday (sorry mods).

Any news on another BETA driver to properly support crossfire and 7970s?

Not yet.

 

[LtMatt]

For me (i7 920 + 2 7970s) the difference in FPS and Smoothness is massive. Even though my cards and crossfire aren't supposedly supported yet.

The only problems I have with 14.1 is a memory leak and huge CPU spikes after a while in game but i'm sure they will get sorted out on a later Beta driver.

For now I'm back on the 13.12 drivers. I'm Really looking forward to the Mantle drivers when they have ironed out the issues.

+1.

 

[LtMatt]

Wonderful demonstration of the Mantle effect. Make sure you watch in 1080P/1440P/2160P quality.

 

[LtMatt]

Looks like Thief will have Mantle from the get go.

 

[LtMatt]

One thign i noticed.

If i start my pc up and do stuff like normal aka check email watch youtube or porn or we and i start BF4 after fps is all over the place CPU usage jumps and ****.

If i reebot my pc and start bf4 str8 after 70-71fps SOLID all the time cpu and gpu usage all normal.

Interesting must be Mantle bug of some srot ??

Driver bug i expect, not a Mantle bug.

 

[LtMatt]

Probably some nice malware from those dodgy porn sites

But Gregorio told me to get the latest Mantle drivers from www.Bangkokchickboys.com

 

[LtMatt]

i'm itching top get a second 290 to add to my water loop, only the look on my wifes face is stopping me atm

I can highly recommend it.

 

[LtMatt]

I had to google what the hell ragnarok2 was. I'm still not sure after looking at pictures of it. Nonetheless id stick with 13.12 WHQL for now until a newer Beta comes out. 14.1 is fine for a teaser of Mantle though.

 

[LtMatt]

haha its a terrible korean game where i play a bear that rides a giant tortoise or sometimes a bike and sidecar

so no good reason to retry 14.1 if i dont play bf?

Sounds a bit like the porno version of animal farm. A true classic.

In short, no not really. Wait for the next beta.

 

[LtMatt]

Good news chaps. Dice are releasing a patch which address some of the problems with Mantle, fixing those lag spikes.

We’ve started rolling out a new Battlefield 4 game update on PC. You may experience some downtime during this rollout, as we are also updating the game servers. As usual, it will take a few hours before the majority of servers are updated.

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

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

Source
http://battlelog.battlefield.com/bf4/forum/threadview/2979150493838365687/

 

[LtMatt]

Fixed.

Get your ass on the server Rofls. All hands on deck.