HollyGamer said:
Random_Matt said:
Ok, just dived deeper and it is copy pasted from the same rumour, AMD Gonzalo. If true, it's 95W TDP is not going to give you that GPU grunt. |
PS4 pro is 330W TDP , Xbox One X is 260W TDP. Original PS4 is 200W TDP .
Also don't forget it will be using 7nm node size and double the amount of CU and shaders on PS4 pro at least and will have 1.35 performance boost with the same TDP
|
This is false.
For doubling of CU's that would mean 36 CU's x2 = 72 CU's.
Graphics Core Next tends to have a hard limit at 64 CU's.
And a doubling of CU's for only a 1.35x performance boost? Ouch. Vega 7 actually has less CU's than Vega 64 and offered a substantial performance increase for roughly the same TDP.
As for TDP's: https://www.anandtech.com/show/10663/analyzing-sonys-playstation-4-pro-announcement
But TDP isn't actually power consumption...
https://www.youtube.com/watch?time_continue=13&v=0wNoCnPxTp4
HollyGamer said:
Random_Matt said:
I do not take rumours seriously, neither should you. You also need to research GCN limitations
|
Of course i am not taking this seriously, i just like tech, PC parts and math like to do some research and make some theory based on the data. GCN limitation will be limited to how much Sony will pay AMD and how big PS5 APU chip size will be.
|
There are some serious technical limitations we need to consider as well.
It's not just about size and cost.
fatslob-:O said:
Now things can change but holy fuck that thing is probably at least $500 I can imagine ...
Still hoping that it'll support hardware accelerated ray tracing as well for a truly next gen leap ... |
Ray Tracing tends to inherently be compute limited, which ironically is one of the big strengths of Graphics Core Next.
I don't think AMD will spend transistors on fixed function blocks for Ray Tracing tasks, but simply make their shader pipelines more flexible to take on the task instead, then as they scale up in shader counts, then performance in old/current games get a boost as well... Should also mean that porting Ray Tracing to older GCN GPU's will likely be a little more feasible.
In saying that, various games have been dabbling in Ray Tracing for years... But RTX Ray Tracing just stepped it up a big level.
Ganoncrotch said:
When the console war was so closely fought during the ps3/360 I think it was one of the most popular times on this site, tons of new users coming here to see who was winning the fight each week, was awesome, now each week people just check in to see if Nintendo or Sony are taking top or bottom while double teaming MS.
|
Not to mention the Wii coming from nowhere and stealing some thunder.
It could happen again, it could not also. Interesting to see how it all plays out.
Bofferbrauer2 said:
Even in 28nm the components shouldn't consume nearly as much power at the clockrates they have I would be very surprised if it goes beyond even just 200W. It's basically a mobile CPU with an underclocked RX 480, which consumes only 150W without that underclocking, and some memory. Where do you got those TDPs?
And I agree with Matt, getting this into something close to 200W would either mean a gargantuan leap for AMD in the GPU department or just a bigger box which allows for more TDP. Right now the GPU alone would consume well over 300W to come near that power. While I think Navi will be less consuming, slashing consumption by over half is just not credible.
Besides, I don't believe the 500$ pricetag for that console, a GPU alone with that power would cost that already. Unless they want to go the PS3 route and subventionize over half the price early on.
The only things credible are the CPU (and even then, not sure about the 16 threads, as games are very hard to parallelize onto multiple threads, more than 8 thus doesn't make much sense in a console as the other threads are useless, so 8/8 would suffice and be safer in light of Spectre), the RAM (with a 384 bit connection) and the HDD (2TB should be minimum by then)
|
Navi is not a Vega replacement though... So anyone expecting Vega levels of performance or TDP probably have their expectations set a little high.
Navi is in-fact a Polaris replacement.
AMD currently flip-flops on high/mainstream GPU releases and iterative GCN updates.
Even the CPU has me guessing. I would expect a single CCX for the CPU side of the equation.
HollyGamer said:
Also Radeon VII also able to beat Vega 64 with the same TDP on half size chip but with higher clock speed.
|
Vega 7 is 331mm2.
Vega 64 is 486mm2.
HollyGamer said:
Don't forget consoles price is not the same with PC GPU card that selling on the market . We also have news RAM price will go down and GPU price also going down up to 30% next year.
|
Correct. But consoles are also at the mercy of PC commodity prices... That is, when DRAM goes up in price... Then that price also increases for consoles.
Conversely... Because Microsoft and Sony are buying fab capacity from the likes of companies such as TSMC... When there isn't much spare capacity, then TSMC can charge an extra premium.
HollyGamer said:
Random_Matt said:
64CU max, how high can you clock a GPU? It would have to be obscenely high I tell ya. |
Just watch Digital foundry , they said it's possible to pass that limit https://youtu.be/5ixSVQwLaZw
|
I would give more credit to Anandtech when it comes to low-level stuff like that.
At the moment, 64 CU's is a GCN hard limit. - Is it entirely impossible to circumvent? Of course not.
But by going past that limit, inefficiencies come into play... And to get around those inefficiencies, some serious re-engineering work needs to be done to Graphics Core Next at a very low level, something AMD is not likely to spend significant time and money on to achieve when their next-generation GPU architecture is coming next year.
I.E. What is possible is mirroring the GPU and using an interconnect for communication. - The chip clusters themselves are still limited to 64CU's though.
It would be no different than getting two Vega 64 GPU's and running them in Crossfire. - You technically have 128 CU's to play with. - But each chip is still 64 CU.
Bofferbrauer2 said:
GCN is limited to 64 CU, so the only way to get Navi faster than that is to clock it faster. To reach 14TFlops almost 2Ghz would be needed, way too much for GCN.
|
1709mhz @ 64 CU's is 14 Teraflops.
Vega 7 will boost to 1750mhz... But because it only has 60 CU's... Means it is hitting 13.8 Teraflops.
But that's not sustained performance... And that is still a 300w TDP part before we even start looking at other components like CPU.
Bofferbrauer2 said:
Add to this 24GB not-yet-released (and thus more expensive, at least early on) GDDR6 Memory (GDDR6 is based on DDR4 btw)
|
You sure?
twintail said:
How does it compare to new xbox rumoured specs? |
Slightly ahead of the more sensible Anaconda specs.
HollyGamer said:
GCN CU are not limited to 64, but Stream processor does (current one is 64 rop's) and it might change (double it in Navi ) or even if it's not they can just added more CU (scalability ).
|
There are 64 Stream processors per Compute Unit. That is a hard limit.
There is a maximum of 64 Compute Units. That is a hard limit.
There is a maximum of 64 Render Output Pipelines. That is a hard limit.
There is a maximum of 256 Texture Mapping units. That is a hard limit.
There is a maximum of 4 Geometry units. That is a hard limit.
There is a reason why Vega doesn't exceed those limits that were hit by Fiji many years earlier.
Navi's main focus will likely be extracting as much as it can from GDDR6 and implementing/improving on a few new tricks introduced with Vega... And hitting the Polaris price point/TDP.
It's a Polaris replacement... And not a high-end GPU, it's main focus is going to be efficiency.
HollyGamer said:
GCN CU are depend on how much chip size they want to put on it while keep it cost and TDP down. Remember with 7nm node it will double down the amount of nodes on the same chip size (PS4 to PS4 pro able to double size the amount of CU)
|
Vega 7 is 7nm and didn't double anything except on the memory side of the equation.
HollyGamer said:
GDDR 6 base price will be different, depend on how much Sony ordered , ( we are talking about a sum size amount of GDDR 6 unit not just a thousand unit) for 5 to 6 years PS5 production . The same thing happen with GDDR5 , it's very expensive but the price is more cheaper because it's a different kind price and deal.
|
Or... They simply might go with GDDR5X or GDDR5 which if taken wide enough can be faster than GDDR6 anyway.
All comes down to price/capacity in the end, it's a balance the console manufacturers will need to weigh.
Although, by the time 2020 rolls around, GDDR6 should be very price competitive even with GDDR5 and certainly have a capacity advantage.
Currently Online 
