NGGP: NextGen Garbage Pile (aka: No one reads the topics or stays on topic) *spawn*

[3dcgi]

He was assuming it was to store the framebuffer. (which would be kinda dumb as it's not really a latency sensitive operation)- As we understand, it's recommended the framebuffer goes to DDR3 ram and shader/compute ops go through the ESRAM.

He wasn't thinking of the big picture.

Framebuffer operations can be latency sensitive if you're blending and need a read modify write operation. If you have enough hardware to hide the anticipated latency it should be fine though.

Does that then mean this kind of extra latency is inherit to consoles with UMAs? Did Xbox/360, or any other console with a single memory pool I may have missed, also have longer latencies compared to PC part equivalents?

Potentially. A unified memory controller will give priority to the CPU because it's less latency tolerant than the GPU. If the CPU is using a lot of bandwidth it would increase the latency relative to a system with dedicated bandwidth for the GPU.

 

[mrcorbo]

I still can't pass DD+, TrueHD and DTS-HD MA to my A/V receiver for decoding on my 60GB PS3. You can't do it on the 80GB or the 40GB PS3s either. I can do it on the Slim and Super Slim PS3s, though. I wonder why that is. According to statements about, that didn't happen.

- 60GB to 80GB removed one hardware feature (GS) and added another (20GB of storage).
- Later models dropped EE as well and added bigger HDD
- bitstreaming was added later with even larger HDDs, later.

BIgger hard drives, when the hard drives are user replaceable and bitstreaming, which has questionable usefulness when you can pass the decoded audio in full fidelity over HDMI don't (to me) equal what was removed. I stand by my statement that the fat PS3 was the best model Sony produced.

 

[arijoytunir]

What is the difference between these 2 metrics ? 1.6 billion triangles/s vs 1.6 Gtri/sec

 

[3dcgi]

What is the difference between these 2 metrics ? 1.6 billion triangles/s vs 1.6 Gtri/sec

Nothing. Different wording for the same thing.

 

[MBDF]

So now the Xcpu has *more* than twice as many flops as the vanilla jaguar. If true, I'm a bit disappointed in Sony. They had always used cpu centric designs in their consoles... I truly hope that the 4CU can manage to make up the difference for their sake.

All in all Durango is looking better and better, not to mention forward thinking. And that hurts to admit, myself being a Sony fan.

 

[Brad Grenz]

Its also important to note that the DMEs in durango, apart from the fact that there are 4 of them, they also perform additional tasks than normal DMAs, eg. each one can tile/untile, one can decode jpeg and LZ while one can encode LZ. These are functions that would traditionally be performed on a cpu or gpu, so it does save some compute resource.

And Orbis is rumored to have its own custom decompression hardware. For all we know it has the same damn DMEs under a different name.

Ok, but doesn't your analysis here also totally ignore other hardware MS is surrounding the GPU with? Can any of those extra parts (eSRAM, display planes, DME's) make up for the 2 CU difference or 200 Gflops gap simply by being significantly more efficient?

No, because we have no reason to believe the custom hardware in the Durango design makes it MORE efficient than Orbis. All evidence suggest they have been added to make sure it isn't far LESS efficient than Orbis.

 

[Silent_Buddha]

So now the Xcpu has *more* than twice as many flops as the vanilla jaguar. If true, I'm a bit disappointed in Sony. They had always used cpu centric designs in their consoles... I truly hope that the 4CU can manage to make up the difference for their sake.

All in all Durango is looking better and better, not to mention forward thinking. And that hurts to admit, myself being a Sony fan.

Stop flogging yourself over rumored specs and the rumored leaks.

The final consoles may perform nothing like what the rumors and leaks thus far suggest. Or they may. Or one might but the other doesn't.

Enjoy the speculation but don't go making final judgements on what you think performance will be like until the consoles are out, and we have more solid information.

This is the time to speculate on what the hardware that is rumored to be in the consoles can do either in a traditional way or something a bit more non-traditional.

This is certainly NOT the time for anyone to be making judgements as to whether console X is faster than console Y.

Regards,
SB

 

[patsu]

The boundary between CPU and GPU is more blur now... oops, I meant nextgen ^_^

 

[astrograd]

No, because we have no reason to believe the custom hardware in the Durango design makes it MORE efficient than Orbis. All evidence suggest they have been added to make sure it isn't far LESS efficient than Orbis.

What evidence? Please lay out the arguments and supporting evidence/logic for me that forces rational thinkers to conclude what you have here. All signs point to MS's specs being settled well prior to Sony's, so I'm not sure how you've convinced yourself that MS reacted to some Orbis specs somehow.

No offense but you sound like you are assuming MS's console was weak a priori and then falling victim to confirmation bias based on that assumption. Why would MS spend more money on designing this setup when the off the shelf option is most likely faster, simpler, and cheaper? Just so they could engineer themselves a performance disadvantage?

Please explain your reasoning for me because your posts here and at GAF come off as very defensive at times and I'd rather not bother wading into some back and forth unless you are willing to share your logic openly. Just saying.

 

[scently]

And Orbis is rumored to have its own custom decompression hardware. For all we know it has the same damn DMEs under a different name.



No, because we have no reason to believe the custom hardware in the Durango design makes it MORE efficient than Orbis. All evidence suggest they have been added to make sure it isn't far LESS efficient than Orbis.

The orbis one is zlib, they are different.
And I don't think they added these things because of orbis. The durango has a performance target and a design phiosophy and so does orbis. These console are for the most part developed on their own. It certainly isn't a "let's see wat sony/ms are doing and then counter them". If you actually think consoles are designed that way, or that parts such as the DMEs/eSRAM or whatever are added at sudden notice then that's a rather naive way of thinking. The info we have one durango and orbis, durango especially are over 10 months old. They most definitely did not add these things to make it far less efficient than orbis. These systems are designed to work the way their makers want them to work and not because of some reactionary whim. The fact that some people think its done this way is not only wrong but I find it rather amusing.

 

[Silent_Buddha]

No, because we have no reason to believe the custom hardware in the Durango design makes it MORE efficient than Orbis. All evidence suggest they have been added to make sure it isn't far LESS efficient than Orbis.

Eh, we already know that the hardware in Orbis is going to be less efficient in some ways than the hardware in Durango.

Just look at the 32 versus 16 ROPs. Those 32 ROPs for example aren't likely going to be fully utilized with the limited bandwidth and CUs that it has compared to something like a Radeon 79xx. The 16 ROPs on the other hand will be able to make better use of it's available bandwidth and resources. If I were to just take a hand waving guess, 20-24 ROPs might be ideal for highest efficiency for the resources available in Orbis. But the ROPs can't be scaled with that fine grained granularity. It's either 16 or 32, nothing in between.

In other words, the 16 ROPs in Durango will have higher utilization (thus higher efficiency) than the 32 ROPs in Orbis in most cases.

Higher utilization does NOT equal higher performance which is where I think you are getting confused. Those 32 ROPs even when limited by the resources available in Orbis are still going to have higher theoretical and likely real world performance compared to Durango.

Regards,
SB

 

[Aeoniss]

Then what is the point of including 32 ROP's? That. Seems like a total waste?

 

[MrFox]

Then what is the point of including 32 ROP's? That. Seems like a total waste?

16 ROPs of Durango will max out at 102GB/s (memory could have allowed more)
32 ROPs of Orbis will max out at 176GB/s (limited by memory)
So the performance of Orbis ROPs will be 72% higher than durango.

None of the AMD cards max out their ROPs, there's always more than enough. Durango is weird, it doesn't have enough ROPs to max out it's 170GB/s. How's that supposed to be called an efficient design?

 

[Bowie]

Both Orbis and Durango should have Z-buffer and Color buffer compression to reduce the bandwidth required by the ROPs.

 

[Aeoniss]

@_@ I see.. Hrm.

So basically you can only increment up to 32 from 16 ROPS? There's no middle ground to be had?

 

[Hecatoncheires]

I was coming back in regards to this as I had some extra info. This does after all seem to be the case for Xbox 3's CPU. I'm not sure on the exact details, but from a FLOPs perspective it seems it may have as much as double the performance of "vanilla" Jaguar cores.

That would either mean Leadbetter, who put his head above the parapet and said he double checked every source and information, is completely wrong, or that both Orbis and Durango use the same 200 GFLOPS enhanced version of Jaguar:

"(...)we've already mentioned that both the next generation Xbox and its PlayStation competitor feature the same CPU - an eight-core AMD offering running at 1.6GHz and based on its forthcoming low-power, high-performance architecture, Jaguar." Digital Foundry

 

[Ketto]

I think Leadbetter (I seriously thought his name was Leadbeater) was saying their the same CPU as in 8 core jaguar chip not so much "they're the exact same CPU".

Potentially. A unified memory controller will give priority to the CPU because it's less latency tolerant than the GPU. If the CPU is using a lot of bandwidth it would increase the latency relative to a system with dedicated bandwidth for the GPU.

That was the case with the original Xbox IIRC(?), which was one of the reasons the GPU never really got to stretch it's legs despite the fact it should have been on par if not better than the Geforce 3 Ti500.

One thing I will say is I love looking at MS' console history, there's a clear progression path that anyone can follow. They build their new console correcting mistakes and adding new ideas they've cooked up since the last one.

 

[Proelite]

That would either mean Leadbetter, who put his head above the parapet and said he double checked every source and information, is completely wrong, or that both Orbis and Durango use the same 200 GFLOPS enhanced version of Jaguar:

"(...)we've already mentioned that both the next generation Xbox and its PlayStation competitor feature the same CPU - an eight-core AMD offering running at 1.6GHz and based on its forthcoming low-power, high-performance architecture, Jaguar." Digital Foundry

LeadBetter was all kinds of off with his Orbis and Durango articles.

 

[Ketto]

Seeing your posts without your usual MLP tank avatar is such a weird feeling

 

[Hecatoncheires]

LeadBetter was all kinds of off with his Orbis and Durango articles.

These articles say basically the same as the articles of EDGE and VGLeaks. So you mean parts of what came to the surface in the last two months are totally wrong and almost every gaming site in the world wide web is chasing a mirage?

I think Leadbetter (I seriously thought his name was Leadbeater) was saying their the same CPU as in 8 core jaguar chip not so much "they're the exact same CPU".

I guess "same CPU" means "same CPU". Why would it mean something else? He's a native speaker of the English language (I'm not btw) so I assume he definitely knows how to put his thoughts to paper.