That we laypeople heard?
Baseless Next Generation Rumors with no Technical Merits [post E3 2019, pre GDC 2020] [XBSX, PS5]
- Thread starter BRiT
- Start date
- Status
That we laypeople heard?
Well, any public leaks... ;P
Makes sense, i highly doubt Sony even cares what MS is doing with their console regarding raw TF performance, in special if it is a two tier launch on MS's side. 8 to 10TF for the PS5, with 8 being the lowest and 10 max, is most likely what's in there. It fits the ballpark 9TF narrative that most sources and leaks point to.
That butterfly structure is familiar.
I am seriously puzzled by the tiny zen 2 cores in those images, surely they need a lot more supporting circuit around them. Otherwise why not put 16 if it's that easy?
Except Microsoft just trade marked power your dreams and midsummer nights dreams was a produced in 1595, 15TF hardware OMG cheeseburger with bacon confirmed!!!
anexanhume
Veteran
The cache is reduced from 32MB to 8MB. Diminishing returns after 8C is my bet.
AbsoluteBeginner
Regular
Perhaps its not as big of a hit in APU as it is on desktop part with chiplet.
That’s for next , next gen
Some of the uncore will be shared with GPU. The memory phy and IO are all going to be shared with the GPU. The cache will likely be severely reduced like they did on Renior. The cores themselves are tiny but I doubt they want to invest in 16 cores when they could add more GPU if they had space/power budget. 8 zen2 cores seems like a good sweet spot to target for next gen, kind of like 8 Jaguars were the sweet spot this gen.
AbsoluteBeginner
Regular
What was expected prediction in frequency increase for iso power going from 16nm to 7nm?
I seem to remember something like this when AMD was announcing 7nm:
Currently TSMC quotes a 20% improve over 10nm and 15% improvement of 10nm over 16nm. If you do the math, they are claiming about 38% improvement on transistor performance.
AMD has also been trying to get higher clocks GPUs even before 7nm. It's very likely that they were targeting higher frequencies on 7nm originally.
https://m.moneydj.com/f1a.aspx?a=909b7f43-d2e8-4cea-9542-690003ab4f9d
Phison denies rumors of PS5 order .
Wrong PCB leaks from reddit:
Phison denies rumors of PS5 order .
Wrong PCB leaks from reddit:
This is not like I have said it before. Every rumor with a standard SSD controller are false. Because it is a custom controller by Sony.
One small caveat: the leak is devkit. It might be a temporary substitute for the WIP Sony controllers.
There is nothing definitively proving the OQA leak right or wrong. I personally don't hold it in high regard.
We don't know if it's custom or standard, right now it's just speculation.
At no point did Cerny excluded a perfectly standard nvme hardware. It could be standard on an nvme socket, and there would be additional hardware in the SoC or the SB to help decryption and decompression at wire speed, except 40 times more power allocated to that asic section than what the ps4 had for LZ and JPG codecs.
I hope that's what they did, so it would be user replaceable as long as you put an nvme pcie 4.0.
It is exotic hardware. We see a minor improvement on PC with nvme and the load times are still significant, while they literally demoed a 0.8s load time on a "slow" ps5 devkit.
I'm just saying the exotic hardware doesn't have to be inside the nvme controller. The SB and SoC are already custom chips, might as well use that development cost to integrate everything they want there instead of a third custom chip preventing upgradeability.
Mostly a software fault. SC does show a larger difference, games need to be optimised for it.
Does this not confirm the Phison controller in Xbox Series X report as being accurate?
Then does that mean the slide from the Samsung SSD Forum is about the PS5? Or just a sales pitch? Or would they provide the SSD but not the controller?
- Status
Similar threads
- Locked
