Predict: The Next Generation Console Tech

Status
Not open for further replies.
A Trinity APU like the AMD A10 5800k delivers around 750 GFLOPS. This is an out of the world performance as compared with the Cell, not least because of the four big Piledrivers contrary to the Cell PPE.

It's all relative though - Trinity offers 3x the max performance of Cell at 6x the transistor count. A modernized Cell would have far outperformed Trinity from a transistor/Flops standpoint, and likely an absolute and wattage standpoint as well.

That's neither here nor there, and as you said, the philosophy at Sony has fundamentally changed as to how the console architectures are approached. Mind you if the previous tales of friction are to be believed, this isn't something that was fully determined or resolved until around 2010 internally.

Anyway, I agree that packaging is going to be where some of the "flavor" is likely to come from Sony's direction this go-around, and I'm onboard with 3dilettante that it in part is likely part of an R&D effort expected to yield dividends across multiple divisions. I know that supposedly, Sony is pushing on 3D imaging sensors right now, and in line with their renewed cell phone push and other IC concerns, it just seems like it would make sense.
 
The media accellerator is used for media processing. The hardware and software used to clean up video file output aren't specced to handle the throughput or perform appropriate work on the output of a 3d game render.

Right thanks for that clarification ..

I was thinking exactly that scenario, that the output from a rendered frame in a scene can be passed through the media for edge detection and then clean up ...

Sounds like that can't happen due to spec .. Did not know that! thx
 
I have a question to those with prior development experience or very familiar with it. I've been trying to think of different scenarios based on info I've received in the past and present. And one is based on a post from Aegies.

Is it plausible that development tools from Sony are not as good as those from MS affecting the possible performance devs get out of PS4? If that doesn't make sense, I'll try to word it better after I get some responses that would give me a better idea of how to improve the question.
 
Is it plausible that development tools from Sony are not as good as those from MS affecting the possible performance devs get out of PS4?

Yes, of course that's possible - there is nothing to clarify or expound upon. This was the entire story of the PS3 at its launch. Since then, however, Sony has put a ton of effort into improving tools, to the point where I would doubt there remains much functional disparity heading into this next generation. Certainly not the gap that existed with last gen at least.
 
Ok. Because being told in the past the power between the two is debatable vs. some of what I'm "seeing" on paper wouldn't seem to indicate it should be as debatable as it's made out to be.
 
You're combining 2 different standards there. The interface is part the standard.
It's either DDR4 or Wide-IO. You can't mix and match them.

Right, my bad. It's one or the other but the principles are similar (stacked DDR4 or Wide IO)

Slide1-2.gif


http://www.infoneedle.com/posting/99670
 
I have a question to those with prior development experience or very familiar with it. I've been trying to think of different scenarios based on info I've received in the past and present. And one is based on a post from Aegies.

Is it plausible that development tools from Sony are not as good as those from MS affecting the possible performance devs get out of PS4? If that doesn't make sense, I'll try to word it better after I get some responses that would give me a better idea of how to improve the question.

absolutely ... ive coded on both PS3 and XBOX360 sdks/tools and MS stack is so much better. PS3 was quite frustrating because there was so much power in the cell BUT it was so hard to access it. So much needed to be done by yourself ..

I'm hoping that PS4 has a much better dev story.

As for MS , its dev story is always great BUT for me its debugging tools were a stand out. I spend around 70% of my time tracing and debugging, and there tools were standout over PS..

I am expecting this generation for both to have a great dev and debugging story!
 
I have a question to those with prior development experience or very familiar with it. I've been trying to think of different scenarios based on info I've received in the past and present. And one is based on a post from Aegies.

Is it plausible that development tools from Sony are not as good as those from MS affecting the possible performance devs get out of PS4? If that doesn't make sense, I'll try to word it better after I get some responses that would give me a better idea of how to improve the question.

Not really when people talk about poor development tools, it's mostly about slow (or in some cases buggy) compilers and linkers and difficult to use or inadequate debuggers.
At the start of the PS3's life MS were also shipping robust profiling tools and Sony really wasn't.

Now though that gap has all but closed, the Sony tools for Vita are embedded in VS, and generally the gold linker used on Sony's existing offerings is faster than the MS equivalent. That speed difference has led a number of people to switch their development platform of choice.
 
Ok. Because being told in the past the power between the two is debatable vs. some of what I'm "seeing" on paper wouldn't seem to indicate it should be as debatable as it's made out to be.

It depends on what you're speaking to here though; what "power" on paper (for the current gen) is indicative of vs the actual real world results? If that's what you mean, some of it is the tools, some of it is dev culture across studios, some of it is performance "left on the table," some is budgets and deadlines, etc etc...

One of the problems with comparing different architectures is that it's not a straight apples-to-apples; there's potential, use-case scenarios, barriers to utilization, all sorts of things. And the tools absolutely play their role there. Which is why this "super Flops" talk recently has all the hallmarks of all the missteps and confusion that followed the 360 and PS3 at their respective launches discussion-wise.

As details emerge, we'll be able to dissect and explain the material differences, advantages, disadvantages. Until then, we can hypothesize to that effect. But people shouldn't let themselves get hung up on one particular aspect of either system as a make-or-break thing, because none of these features will live in isolation. The systems need to be viewed holistically.
 
NAt the start of the PS3's life MS were also shipping robust profiling tools and Sony really wasn't.
The profiling tools should make a big difference, no? That's what I was thinking. If you are hitting a bottleneck in your executable and it's hard to trace and there's a deadline coming, you'd just let it slide. Hence poor dev tools leading to lower hardware utilisation.
 
Ok. Because being told in the past the power between the two is debatable vs. some of what I'm "seeing" on paper wouldn't seem to indicate it should be as debatable as it's made out to be.

So the 1.8 and 1.2 numbers are close to right and the secret sauce isn't some wunderkind. That's what made sense so I'm not surprised to hear it.
 
I have a question to those with prior development experience or very familiar with it. I've been trying to think of different scenarios based on info I've received in the past and present. And one is based on a post from Aegies.

Is it plausible that development tools from Sony are not as good as those from MS affecting the possible performance devs get out of PS4? If that doesn't make sense, I'll try to word it better after I get some responses that would give me a better idea of how to improve the question.


BTW this generation of consoles and the sdks/tooling for me will need to have a great Parallelism and HSA story. I know how MS and it's tooling around both of these is progressing, I know next to nothing of how Sony will provide me these in there tooling..
 
It's all relative though - Trinity offers 3x the max performance of Cell at 6x the transistor count. A modernized Cell would have far outperformed Trinity from a transistor/Flops standpoint, and likely an absolute and wattage standpoint as well.

That's neither here nor there, and as you said, the philosophy at Sony has fundamentally changed as to how the console architectures are approached. Mind you if the previous tales of friction are to be believed, this isn't something that was fully determined or resolved until around 2010 internally.

Anyway, I agree that packaging is going to be where some of the "flavor" is likely to come from Sony's direction this go-around, and I'm onboard with 3dilettante that it in part is likely part of an R&D effort expected to yield dividends across multiple divisions. I know that supposedly, Sony is pushing on 3D imaging sensors right now, and in line with their renewed cell phone push and other IC concerns, it just seems like it would make sense.

Presumably if they go this route, it's easier to support b/c and cross platform development moving forward (downport from PS4 and Vita + other mobile platforms ?)

I'd like to see how PhryeEngine and other commercial ones can help here.
 
The profiling tools should make a big difference, no? That's what I was thinking. If you are hitting a bottleneck in your executable and it's hard to trace and there's a deadline coming, you'd just let it slide. Hence poor dev tools leading to lower hardware utilisation.

I think ERP was speaking to PS4's potential handicaps (or lack thereof) when he started with "not really" though, and then using PS3's origins as the backstory for the continued perceptions.
 
One of the problems with comparing different architectures is that it's not a straight apples-to-apples;

As details emerge, we'll be able to dissect and explain the material differences, advantages, disadvantages.
I'm assuming bgassassin's looking at paper specs that have significant architectural similarities, hence on paper Orbis should beat Durango. Let's say his sources have told him both are 8 core Jaguars at 2 GHz, 8800 type GPUs, one at 1.2 TFlops and one at 1.8 TFlop (apple to apple comparison), one with 100 GB/s BW and one with 200 GB/s BW. That would point to the second machine being the more powerful, and so comments that there's not much between the machines would be hard to reconcile. At which point, you'd want an explanation, and bg's asking if it could be dev tools. I don't think he'd ask his question if there wasn't good reason to see the hardwares as comparable. If the answer to his question is 'no', then he knows either the development comments are wrong, or the hardware specs he's seeing are wrong, or there's more to the machines than the specs are saying.
 
Presumably if they go this route, it's easier to support b/c and cross platform development moving forward (downport from PS4 and Vita ?)

I'd like to see how PhryeEngine and other commercial ones can help here.

Well, b/c going forward for sure - but that's two "gaps" in the legacy support now. In any event, ISA completely aside, what I am really looking to begin discussion on is what terms Sony received in this partnership with AMD. Everyone is familiar with the contractual disaster that was Intel/NVidia for MS the first go around, so doubtless this will be structured quite differently. Not to mention, AMD is in life support mode these days, and I would hope that to the extent things are "off the shelf" as such, that the sourcing costs reflect lifeline pricing.

This is definitely my #1 key point of intrigue insofar as Sony's decision to go AMD was concerned. Certainly keeping costs down is going to be priorities #1, #2, and #5.

PS - And yes, definitely interested to see how PhyreEngine and its evolution will play a role. I would think it will be a cornerstone of the overall dev environment.
 
Status
Not open for further replies.
Back
Top