superchunk said:
I doubt more power is somehow being added. Its far more likely that Naughty Dog, being a quality game developer studio, is making a superb game that demonstrates how you can utilize the PS4 SDK to garner better results.
Then they are giving away their tweaks that will likely become an SDK update to PS4 for everyone else. |
Yep ! ICE Team Playstation 4 Performance Boost Thanks to GPGPU Rendering
Much like Microsoft is trying to wring every last drop of performance out of the Xbox One, Sony must do the same for the Playstation 4. A few days ago (May 19th) rumors surfaced that the PS4 would be on the receiving end of a ‘massive performance boost’ after Tidux brought up the subject on Twitter. We’ve speculated as to how this could be accomplished, and as it turns out it’s got something to do with the SDK of the PS4 and GPGPU rendering.
Originally there were several theories being banded about, but industry insider Ashan Rasheed (perhaps better known as Thuway) replying to someone on Twitter who was querying if the performance boost would be something Sony would announced . “Thats more to do with SDK’s improving. ICE Team is legendary with their work on the Sony tools side. You won’t hear about it,” started Thuway.
“Future SDK improvements on PS4 will focus on GPGPU rendering. This is when delta between PS4 and X1 will magnify,” he said.
Truthfully the fact this is SDK related isn’t surprising given SDK’s and toolsets are what ICE Team traditionally work on. Given how early we are in this generation (particularly since Project Morpheus isn’t released yet) it doesn’t make sense for Sony to start playing around with the PS4′s RAM reserves. This is truer still when you further consider than in just seven months there’s no way the PS4 could be close to having its potential fully tapped into.
Mark Cerny (the lead architect of the PS4) had said all along during several interviews the PS4′s GPGPU (compute) capacity would be the key to the consoles performance and this only further enforces his point and philosophy.
For those who’re unfamiliar with the PS4′s basic GPU architecture, it features 18 compute units, housing 64 shaders (ALU) each. This provides a grand total of 1152 processors capable of being used in a variety of different ways. The PS4′s GPU was expanded similar to AMD’s own Volcanic island desktop / desktop GPU line up by beefing up the ACE (Asynchronous Compute Engines) to eight from the previous two.
The Asynchronous Compute Engines queue and control the GPU’s task scheduling, effectively ensuring that compute data is being synced up to the current version in memory. Another important task they perform is the scheduling of when a task is performed so that it doesn’t interfere or drain away processing power from graphical tasks.
http://www.redgamingtech.com/playsta...-similarities/Playstation 4 GPU & Next Gen AMD Radeon Volcanic Island GPU Compute SimilaritiesThe Playstation 4 and the Radeon R9 280X(known as AMD Volcanic Islands) may share quite a few similarities with each other, in particular in the way their Asynchronous Compute Engines (ACE) function. Both the Playstation 4 and AMD’s soon to be released Volcanic Island GPU feature 8 ACE, rather than the 2 featured in the vanilla Sea Island GPU (which is effectively what the PS4 and Xbox One GPU’s are built around).
Both GPU’s ACE units can 8 compute queues, and so total that gives 64 compute commands (using the math of 8 compute queues * 8 ACE = 64 compute commands). We know that Sony believes that compute is going to be incredibly important in the next generation of graphics, with the basic premise being that CPU’s aren’t well suited to processing certain tasks. The reason behind this is that a CPU core can only process one command at a time. Modern day GPU’s on the other hand work in a SIMD (Single Instruction Multi Data) fashion. Hundreds (or in some cases, thousands) of processors (known as SP’s – Stream Processors for AMD’s GPU, or CUDA cores for Nvidia’s GPU’s) all work together to process a piece of data.What are the PS4′s / AMD Asynchronous Compute Engine’s and how do they work?The GCN architecture (Graphic Core Next) uses 2 ACE units, which could handle 2 compute queues (giving a total of 4. 2 Asynchronous Compute Engines * 2 compute queues = 4). The GCN architecture was used for AMD’s previous generation cards, including the AMD Radeon HD 7970 and the later Radeon R9 280X.- Sony had worked very closely with AMD with the desire to beef this up for the Playstation 4, with the intention of offloading much of the work from the CPU to the GPU. Let’s start by explaining how all of this works:
The Asynchronous Compute Engine basically tells data were to go, queuing up the commands and then processing them when the GPU has the spare “Cycles” to process it.
Imagine yourself driving on a highway, with hundreds of lines – but ahead of you is a toll booth. Consider that you can’t see which booths are free from your perspective (you’re stuck behind large trucks and so on after all). So you’re effectively only able to use signs tell you which lane to take. The more ACE’s there are (in this case, electronic signs) to tell you which lane to go to, the faster you’ll be sent to a lane that’s tollbooth is as empty (or at least as free) as possible...
In addition to the increased number of ACE engines, they can queue more instructions (also raised from two to eight) meaning there’s a huge increase in the GPGPU control structure and performance inside the PS4. Shortly put, it can store more compute commands and has a larger number of units to tell compute (or graphic commands) how to run over the shaders. Don’t forget the PS4 also has more GPU performance than the X1 too.
Rounding all of this off is the PS4′s advanced HUMA memory architecture and of course the volatile bit compute structure which allows selective lines of code in the level 2 cache of the GPU to be flushed rather than needing to perform an entire Level 2 flush that would otherwise be necessary.
So, with ICE Team improving the SDK to allow games developers better libraries and an easier time of getting up and running we should see an advancement of GPU Rendering. GPU rendering is a fairly large field and could many a plethora of different tasks. Everything from Ray Tracing to having the GPU calculate certain types of physics or to help in AI and overall production of a scene in the game.
http://www.redgamingtech.com/ice-tea...gpu-rendering/
For comparison :
The XBone GCN architecture uses 2 ACE units, which could handle 8 compute queues (giving a total of 16! --> 2 Asynchronous Compute Engines * 8 compute queues = 16). The GCN architecture was used for AMD’s previous generation cards, including the AMD Radeon HD 7970 and the later Radeon R9 280X.
Sony had worked very closely with AMD with the desire to beef this up for the Playstation 4, with the intention of offloading much of the work from the CPU to the GPU. ---> The Playstation 4 and the Radeon R9 280X(known as AMD Volcanic Islands) may share quite a few similarities with each other, in particular in the way their Asynchronous Compute Engines (ACE) function. Both Playstation 4 and AMD’s soon to be released Volcanic Island GPU feature 8 ACE, rather than the 2 featured in the vanilla Sea Island GPU (which is effectively what the PS4 and Xbox One GPU’s are built around).
Both GPU’s ACE units can 8 compute queues, and so total that gives 64 compute commands (using the math of 8 compute queues * 8 ACE = 64 compute commands).
Even though both PS4 and XBone are built over similar architecture, PS4 not only has 6 additional CU for more power and heavier GPGPU, but also has these 64 compute commands VS only 16 on XBone(I'm pretty sure about it).
PS:
The Asynchronous Compute Engine basically tells data were to go, queuing up the commands and then processing them when the GPU has the spare “Cycles” to process it.
Imagine yourself driving on a highway, with hundreds of lines – but ahead of you is a toll booth. Consider that you can’t see which booths are free from your perspective (you’re stuck behind large trucks and so on after all). So you’re effectively only able to use signs tell you which lane to take. The more ACE’s there are (in this case, electronic signs) to tell you which lane to go to, the faster you’ll be sent to a lane that’s tollbooth is as empty (or at least as free) as possible...
From the link above. Very interesting. I'm confident we will see huge improvements in the following years.
------------------------------------------------------
Confirmed :
Finally, the Xbox One features 2 ACE, each handling 8 queues (16 total). PS4 features 8 ACE, each handling 8 queues (64 total).ACE (or asynchronous Compute Engines) are what handle issuing GPU compute operations to the shaders. Quite simply put, they store a command (in queues) and then the ACE’s tell the GPU’s shaders how it should process it, and how many shaders can do it. Efficient use of these Asynchronous Compute engines is vital, since the commands must be ran while the GPU is not processing graphics to avoid impacting performance of graphics.
http://www.redgamingtech.com/xbox-on...-architecture/
----------------------------------------------
Uncharted4 will be the first example of serious improvements !
”Every great dream begins with a dreamer. Always remember, you have within you the strength, the patience, and the passion to reach for the stars to change the world.”
Harriet Tubman.
