[EG] AMD CPU's performance get "massive boost" as devs optimize next-gen game engines

Page 6 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

OCGuy

Lifer
Jul 12, 2000
27,227
36
91
The Xbox 360 was ahead of its time with a tri-core processor already. I'm not sure the specifics of how well it balanced the load.
 

maddie

Diamond Member
Jul 18, 2010
4,787
4,771
136
I don't think they would have asked for something that would make programming harder like heavily multithreading. Sony and MS losses with hardware in the last round of consoles should have spoke louder than any developer. AMD's razor thin margins too.
I would imagine, seeing that cpu speed vs power consumed is not linear and that,
1] The power budget was in a small range
2] The production cost budget was also constrained

They might have been given a range of solutions. A few higher speed cores or much more lower speed ones. It is now seems the one taken might be the best one, in their minds, but I'm no expert, and probably few of us here can say with complete certainty.
To say is was basically forced on them as a fact is without evidence.
 

StrangerGuy

Diamond Member
May 9, 2004
8,443
124
106
I am looking at today. What I see are tons of titles optimized for consoles, which run okay on PC simply because the PC has brute force power even if the game is very poorly optimized (from the point of view of the PC).

Why do you assume that in the future, games won't be optimized for consoles?

It seems like an absurd point of view based on a long history of console games being optimized for consoles.

Some people just don't get it though. As much as graphically revolutionary Crysis was back in 2007 it only sold 1M copies. Current development costs are already so high on current gen that to maximise today's PC hardware just to sell PC games is next to impossible.
 

Dribble

Platinum Member
Aug 9, 2005
2,076
611
136
I think people here are over-estimating how hard multi-threading is. Been doing it in my job for many years. It's really not that difficult - you just need the low level libraries to allow you to access data and pass messages around in a thread safe way, then you pick the correct things to thread (renderer, physics, collision) and do it.

These next gen consoles with their standard x86 architectures and in Sony's case even simpler with a unified memory architecture should be easy to code for.
 

Broheim

Diamond Member
Feb 17, 2011
4,592
2
81
I think people here are over-estimating how hard multi-threading is. Been doing it in my job for many years. It's really not that difficult - you just need the low level libraries to allow you to access data and pass messages around in a thread safe way, then you pick the correct things to thread (renderer, physics, collision) and do it.

These next gen consoles with their standard x86 architectures and in Sony's case even simpler with a unified memory architecture should be easy to code for.

I remember writing a small multiplayer p2p arcade shooter in school, multithreading the core code of the game probably complicated things tenfold. I never got it to work properly, 1-3 players were fine but after that things started getting weird. that was a pretty simple little game and I can only imagine how much more complicated things get with big impressive games like ps2. Features exponentially add complexity and complexity adds cost.
 

Aikouka

Lifer
Nov 27, 2001
30,383
912
126
I think people here are over-estimating how hard multi-threading is. Been doing it in my job for many years. It's really not that difficult - you just need the low level libraries to allow you to access data and pass messages around in a thread safe way, then you pick the correct things to thread (renderer, physics, collision) and do it.

Don't collisions normally require physics calculations and all of those are prerequisites for rendering? I guess I'm trying to see how you are necessarily breaking off parallel tasks when they still appear to have a dependency-based ordering.
 

Chiropteran

Diamond Member
Nov 14, 2003
9,811
110
106
Some people just don't get it though. As much as graphically revolutionary Crysis was back in 2007 it only sold 1M copies. Current development costs are already so high on current gen that to maximise today's PC hardware just to sell PC games is next to impossible.

Yeah, people don't get it. Those absurdly high dev costs are not high because the programming is so expensive, they are high because modern game companies are spending the vast majority of the development money on marketing and art (graphics).

http://mwomercs.com/forums/topic/109179-infographic-video-game-budget-breakdown/

Programming is a small part as is. Even if they double the money put into programming, it doesn't necessarily result in higher overall development costs- they can just cut some graphics and/or marketing to keep costs down.

Notice the bit about outsourcing. They can hire some kid in India or China to write the code for pennies an hour. It might be expensive to hire spoiled American programmers, but it's not that expensive if you go with the lowest bidder
 

Dribble

Platinum Member
Aug 9, 2005
2,076
611
136
I remember writing a small multiplayer p2p arcade shooter in school, multithreading the core code of the game probably complicated things tenfold. I never got it to work properly, 1-3 players were fine but after that things started getting weird. that was a pretty simple little game and I can only imagine how much more complicated things get with big impressive games like ps2. Features exponentially add complexity and complexity adds cost.

That's probably a combination of not having the correct underlying libraries to allow you to do threading safely, and not really knowing what you were doing so not coding it in a way suited to threading. Still well done for having a go at it

Don't collisions normally require physics calculations and all of those are prerequisites for rendering? I guess I'm trying to see how you are necessarily breaking off parallel tasks when they still appear to have a dependency-based ordering.

Well you would probably be rendering the last frame while working out what's happening in the next one. Even if you *had* to do them one after another most of those sort of tasks (rendering, collision, physics) would parallelise to 8 threads easily. Most of them are just working though a bunch of data that doesn't change (so no write locks) repeatedly doing a lot of the same maths.
 

Aikouka

Lifer
Nov 27, 2001
30,383
912
126
Well you would probably be rendering the last frame while working out what's happening in the next one. Even if you *had* to do them one after another most of those sort of tasks (rendering, collision, physics) would parallelise to 8 threads easily. Most of them are just working though a bunch of data that doesn't change (so no write locks) repeatedly doing a lot of the same maths.

Ah, I think I see where my train of thought went askew. For some reason, I kept thinking rather procedurally; in other words, if you detect a collision, you gotta work on it immediately. Although, I doubt there would be an issue just tossing the physics request at a worker thread when its encountered and going along your merry way at whatever else is on your queue. As long as you can process it fast enough, the end-user shouldn't even notice, and since consoles are a hard platform, that should be much easier to tweak.
 

blackened23

Diamond Member
Jul 26, 2011
8,548
2
0
I think people here are over-estimating how hard multi-threading is. Been doing it in my job for many years. It's really not that difficult - you just need the low level libraries to allow you to access data and pass messages around in a thread safe way, then you pick the correct things to thread (renderer, physics, collision) and do it.

These next gen consoles with their standard x86 architectures and in Sony's case even simpler with a unified memory architecture should be easy to code for.

Sony and MS are also both providing significant tools in order to program directly to the hardware which makes such tasks (multi threading) far easier, from what i've read.

In fact, I recall Carmack speaking at Quakecon about this very subject. He spoke very highly of the tools provided by both Sony and MS and he seemed confident that the multiple cores on the next-gen systems will provide a significant net benefit. Now I have some grievances with the games that id has put out recently(I *really* hated Rage), but I won't deny the genius of Carmack based on his old works (doom, quake, etc) - if he's saying this stuff, I would take it at face value. Console developers absolutely will optimize heavily for multithreaded/core in every way imaginable. What remains to be seen is whether that translates with PC ports of the same games - that isn't a certain outcome because the next-generation consoles are not using D3D or DirectX. They're using system specific tools to program direct to hardware, which by definition avoids APIs altogether. So we'll have to wait and see whether PC ports benefit.

I just find it hard to believe that some folks think that consoles won't be heavily optimized for - the performance mustered from the 10 year old xbox 360 is pretty unbelievable. They are *still* finding ways to further optimize games for prior generation consoles, and they have squeezed more out of those systems than should be possible. The next-generation systems will be much the same, developers will heavily optimize for them because their livelihoods depend on it. Console games = money. As I mentioned earlier, though, PC ports may or may not benefit. We'll have to wait and see.
 
Last edited:

Obsoleet

Platinum Member
Oct 2, 2007
2,181
1
0
AMD players who've been having sub-par performance on the PC will suddenly get a massive boost - just because of being able to take the engine and re-implement it as multi-threaded.

^ This. It force multithreading that is ideal for PCs that have similarly designed AMD CPUs (with enough distributed thread loading it may fully negate Intel's additional performance-per-core advantage).

Many anti-AMD fanboys are poo pooing on AMD's grand slam, yet an admitted (and extremely knowledgeable) Intel+Nvidia fanboy with a long history buddying up to Nvidia- John Carmack is honest enough to say similar comment in his keynote this year that this is excellent for AMD. http://www.youtube.com/watch?v=eNWAcEu1jpU#t=16m30s

My advice, next time you're buying, buy your AMD APU, CPU, GPU. Take your pick but I'd make it AMD. If you disagree, I'm not taking complaints at this time- talk to Carmack.

It might take a while to pay off dividends, but the recent console design wins by AMD sealed the deal for my next rig (whenever that time comes), which will be all AMD.
 

videogames101

Diamond Member
Aug 24, 2005
6,777
19
81
^ This. It force multithreading that is ideal for PCs that have similarly designed AMD CPUs (with enough distributed thread loading it may fully negate Intel's additional performance-per-core advantage).

Many anti-AMD fanboys are poo pooing on AMD's grand slam, yet an admitted (and extremely knowledgeable) Intel+Nvidia fanboy with a long history buddying up to Nvidia- John Carmack is honest enough to say similar comment in his keynote this year that this is excellent for AMD. http://www.youtube.com/watch?v=eNWAcEu1jpU#t=16m30s

My advice, next time you're buying, buy your AMD APU, CPU, GPU. Take your pick but I'd make it AMD. If you disagree, I'm not taking complaints at this time- talk to Carmack.

It might take a while to pay off dividends, but the recent console design wins by AMD sealed the deal for my next rig (whenever that time comes), which will be all AMD.

Stop with this bullshit. Even if a game is optimized for multi-threading, Intel cpu's will the the SAME benefit as amd cpu's, this whole thread is pointless..

Because 8 integer cores are just going to kill it for game engine computations. :hmm:

just because multithreaded scenarios are the only situation amd cpu's don't suck doesn't mean intel cpu's suck at "multi-threading"
 
Last edited:

StrangerGuy

Diamond Member
May 9, 2004
8,443
124
106
Stop with this bullshit. Even if a game is optimized for multi-threading, Intel cpu's will the the SAME benefit as amd cpu's, this whole thread is pointless..

Because 8 integer cores are just going to kill it for game engine computations. :hmm:

When AMD gets framerates well above silky smooth gameplay Intel can get a billion FPS and it still won't matter.
 

Obsoleet

Platinum Member
Oct 2, 2007
2,181
1
0
Stop with this bullshit. Even if a game is optimized for multi-threading, Intel cpu's will the the SAME benefit as amd cpu's, this whole thread is pointless..

You don't comprehend what you read. I said negate, and it was in bold to prevent daft responses like yours- hopefully.
Negate : to cause to be ineffective or invalid

Intel won't get any benefit unless the AMD optimizations (which will be there) overload the PS4/XB1 APU.
And I'll answer that question you have right now: it will not overload your i7 cores nor grant Intel a continued advantage.

They will likely end up performing the same. While that's a really good estimation from all available evidence. It is still possible AMD might not just negate Intel's lead but could come out with a performance advantage. Only AMD really knows the answer to that today, until we get our hands on an engine optimized for AMD's architecture.
 

beginner99

Diamond Member
Jun 2, 2009
5,223
1,598
136
Notice the bit about outsourcing. They can hire some kid in India or China to write the code for pennies an hour. It might be expensive to hire spoiled American programmers, but it's not that expensive if you go with the lowest bidder

Yeah outsource to clueless sweatshops in India. If your lucky the program will just not run and throw errors but it is far more likely it will blow up your computer. Joke aside. In programing you get what you pay for. And this means if you pay little you will get crap but if you pay a lot you may also get crap (keyword high paid consultants).

Or "If you think experts are expensive wait what amateurs will cost you."

EDIT:

lol, and that picture in your link is obviously a joke and not serious.


Racism and negative stereotypes are simply not acceptable.

Regarding: "Yeah outsource to clueless sweatshops in India."

There are clueless sweatshops in all countries and all races. And there are top-notch sweatshops to be found in all countries as well.

Please do not negatively characterize any given nation, race, or ethnicity. There are far more productive and socially acceptable ways to communicate your key points.

Administrator Idontcare
 
Last edited by a moderator:

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
You don't comprehend what you read. I said negate, and it was in bold to prevent daft responses like yours- hopefully.
Negate : to cause to be ineffective or invalid

Intel won't get any benefit unless the AMD optimizations (which will be there) overload the PS4/XB1 APU.
And I'll answer that question you have right now: it will not overload your i7 cores nor grant Intel a continued advantage.

They will likely end up performing the same. While that's a really good estimation from all available evidence. It is still possible AMD might not just negate Intel's lead but could come out with a performance advantage. Only AMD really knows the answer to that today, until we get our hands on an engine optimized for AMD's architecture.

How did you come to that conclusion? It will first of all be optimzied for Jaguar uarch and the tight timings in the console CPUs. That by default leaves the FX series out. (And the entire PC for that matter.)

Secondly, there is no way any optimizations would get anywhere the 50%+ IPC lead Intel got. Meaning the 4670 would still be the much better gaming chip in all cases since consoles will max use 6 cores for gaming. The only single time that AMD can compete is, when it uses 8 cores and the scaling is essentially 100%. Games dont scale 100% either.

Its simply another hype that this will change anything.
 

inf64

Diamond Member
Mar 11, 2011
3,765
4,223
136
So wait, now intel has +50% advantage in gaming ?? Lmao
Unless you game at 480x240 it's safe to say that in single GPU scenarios an FX is nowhere near 50% slower than any intel chip (even in those notorious intel loving games like SC2).

Take a look how "lowly" 6800K performs in games with high end GPU in 19x12(the resolution people use when they have high end GPU like GHZ Edition 7970). It's indistinguishable from 3570K in 19x12 in BF3, TomRaider and FC3. I guess SC2 and few other games would show a lead for intel but 3570K is much more pricier than 6800K and the average gaming performance in single GPU scenario in 19x12 is very comparable.
 
Last edited:

StrangerGuy

Diamond Member
May 9, 2004
8,443
124
106
So wait, now intel has +50% advantage in gaming ?? Lmao
Unless you game at 480x240 it's safe to say that in single GPU scenarios an FX is nowhere near 50% slower than any intel chip (even in those notorious intel loving games like SC2).

If Blizz can make a D3 port for a 8-year old console it's not hard to see how sloppy they got making for the stutterfest PC version...
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
145
106
So wait, now intel has +50% advantage in gaming ?? Lmao
Unless you game at 480x240 it's safe to say that in single GPU scenarios an FX is nowhere near 50% slower than any intel chip (even in those notorious intel loving games like SC2).

In SC2 Intel is actually 100% faster...



LMAO??? :awe:
 

inf64

Diamond Member
Mar 11, 2011
3,765
4,223
136
Yup, game is just a turd unfortunately . 70$ (+46%) pricier i5 it is then if you play SC2 . In vast majority of other cases 1-2 fps difference. Seems like a great deal :awe:
 

inf64

Diamond Member
Mar 11, 2011
3,765
4,223
136
Also note how pathetic the fps is even on i5, I would hardly classify that as "playable". But hey 11fps is much moar than 7fps in min. department, so much you can have a tad better slide show :awe:
 
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/    |