I think he said that more like you choose to use like you want (14+4, 12+6, 16+2, etc).

Kinda sorta, it was designed the 14 pairs of CU's pair off with the CPU, which leaves an optional 4 left that the developers can use in different ways.  Odd note: the Xbox One's GPU is also made with 14 CU's, but they only use 12 of them, because they calculated they might lose one or two because of chip yield. Intel has been doing that sorts of chip manufacturing style for years.

As I understand, the only "audio" the GPGPU PS4 may do is raycasting audio where it needs to know the 3D render models positions.

The GPGPU will NOT be used to do audio sound effects like audio mixing and such, it would be a wholly inefficient job. The CPU will have to do those effects. Anyway in most cases those effects are really not much demanding on the CPU and most of the time it is the question of the volume of the audio stream.

For raycasting audio the XOne will need to use its own GPGPU. Apparently the XOne would also need to use its Jaguar for some effects like reverbs which are not done in Shape.

I think some people should maybe believe more what PS4 architect is saying than what MS spin machine is telling them when it comes to PS4. Ethomaz already posted Cerny's words, I really don't understand where this "PS4 is designed with 14 CUs in mind for graphics, just like XOne" comes from.

Reminder:

PS4s GPU is 1152:72:32 config with 176GB/s bandwidth (18CUs), 7850 is 1024:64:32 with 153,6GB/s (16CUs), 7870 is 1280:80:32 with 153.6GB/s (20 CUs)

XOne is 768:48:16 (12CUs), 7770 is 640:40:16 (10CUs) and 7790 is 896:56:16 (14CUs)


There is no "4 are customized" for general tasks, if you need them all for gfx, you use them all in that way. PS4 has additional customization to enable compute simultaneously with gfx tasks.

"Next, to support the case where you want to use the GPU L2 cache simultaneously for both graphics processing and asynchronous compute, we have added a bit in the tags of the cache lines, we call it the 'volatile' bit. You can then selectively mark all accesses by compute as 'volatile,' and when it's time for compute to read from system memory, it can invalidate, selectively, the lines it uses in the L2. When it comes time to write back the results, it can write back selectively the lines that it uses. This innovation allows compute to use the GPU L2 cache and perform the required operations without significantly impacting the graphics operations going on at the same time -- in other words, it radically reduces the overhead of running compute and graphics together on the GPU."

"Our overall approach was to put in a very large number of controls about how to mix compute and graphics, and let the development community figure out which ones they want to use when they get around to the point where they're doing a lot of asynchronous compute."

Cerny expects developers to run middleware -- such as physics, for example -- on the GPU. Using the system he describes above, you can run at peak efficiency, he said.

"If you look at the portion of the GPU available to compute throughout the frame, it varies dramatically from instant to instant. For example, something like opaque shadow map rendering doesn't even use a pixel shader, it’s entirely done by vertex shaders and the rasterization hardware -- so graphics aren't using most of the 1.8 teraflops of ALU available in the CUs. Times like that during the game frame are an opportunity to say, 'Okay, all that compute you wanted to do, turn it up to 11 now.'"

http://www.gamasutra.com/view/feature/191007/

Where is your source that developers can take up to 4 CU's of the GPU for sound processing? Or are you just stating your own opnion like the more significantly powerful CPU and co processors from the other article?  The PS3 audio chip is absolutley fine for audio on the PS3 and thats 8 year old technology. By all accounts the PS4 chip is better so no problems there. This co processors secret sauce PR rubbish from MS is just that rubbish. MS have just gave fancy names to chips that are also in the PS4. There are no amazing co processors that are gonna change the power advantage. No amount of spin will change that. Gonna be some major meltdowns come release time.

But I only use .flac and .midi for audio playback, does that mean the PS4 won't do my music (huehue)

That is complete and utter rubbish.

There are 20 CUs in the PS4 gpu. 2 CUs are there for redundancy (just like in the XBox One which has 14 CUs, also 2 for redundancy). These 20 CU units are completely identical in their functions down to the last transistor. The whole point of redundancy would be pointless if there were any "special" CUs present.

Did you read the artlicle from Mark Cerney?  You may want to look at it again if you skimmed it.

You sound upset, are you ok?  I'm not even going to debate the PS3 audio issue with you. They are well documented, but that's all last gen.

Also, I do agree with you about the major meldowns; however, I think we may have different opinions about where they will fall.

All chip manufacturers and major console players have been doing this for years. The Cell in PS3 was designed with an extra SPE but was built in for redundancy. Most GPU cards that fail to meet the manufacturing specs can be re-purposed and sold as lower end cards etc.

No problem, it's just that I usually see you really defensive of Xbox, but credit must be given when it should... and xbox audio processor seems to be stronger, we aren't sure on how much it will be for game and how much for kinect and stuff (that I don't care much, but certainly there will be demand or at least a fad - for some)