_rogame thinks Scarlett die size could be ~342 mm², which is smaller than Scorpio:
My name is Chris and I lurk quietly waiting for answers.This thread is basically a support group.
My name is Chris, and it's been 0 minutes since my last relapse.
So he brought attention to outdated info on the PS5 and now he's potentially giving wrong info on the Series X lol.Completely unfounded. They are scaling off a custom metal carrier instead of an industry standard capacitor size.
Now assuming Phil did not use a lens with too wonky a perspective (eg: some fisheye lenses), our results show us that the Xbox Scorpio and Xbox Scarlett APUs are roughly the same in width but the Xbox Scarlett APU is slightly longer in height. Our calculations put the Xbox Scarlett APU at roughly 401mm², which is an increase of almost 42mm² over the Xbox Scorpio chip at 359mm². The brand new APU features hardware-based raytracing (which seems to be all the rage these days) and a Zen 2-based CPU core.
Scorpio had 2560 SPs with 2 CUs disabled and while the Xbox Scarlett APU is clearly bigger, it is not possible to speculate based just on the die size because of the different process node involved (7nm vs 16nm) but we can safely say that you are looking at at least 50% more power in the same die space. Since the Scorpio APU can output roughly 6 TFLOPs of power, we can guesstimate a range of around 9 TFLOPs for the Scarlett APU.
I don't see any error, but 540mm² is just not happening. That's a 50% bigger dieI counted the pixels high of Scorpio which was 126 and a length of 185
Scarlett is 161 high and length of 218
15.45x23.175 = 358mm2 roughly (i just did some quick maths to get 359mm2 at 1:1.5 ratio and settled at 358 just because i couldn't be arsed to fine tune it)
dividing scorpio 15.45 by 126 and multiplying by 161 gets me 19.74
dividing scorpio 23.175 by 185 and multiplying by 218 gets me 27.31
19.74 x 27.31 = 539
did i do something wrong in there?
The capacitors are constant?Wait he is using the bracket to measure the chip? 😂 Why not use the capacitors...
Using the bracket to measure the chip is like using the console size to measure the die size.
its not that serious, hes just speculatingSo he brought attention to outdated info on the PS5 and now he's potentially giving wrong info on the Series X lol.
Completely unfounded. They are scaling off a custom metal carrier instead of an industry standard capacitor size.
I know it's not serious, I think it's funny lol.its not that serious, hes just speculating
and did you not read literally the first few words of his tweet
Yeah I know the obvious takes incoming but....
only 9TF?!Usman Pirzada measurements:
Yeap. There is more as well if you dig.So he brought attention to outdated info on the PS5 and now he's potentially giving wrong info on the Series X lol.
surface mount caps come in standard sizes, yes. 0402, 0603, 0201, 01005, etc.
Guessing these are 0612 given the aspect ratio.Yeap. There is more as well if you dig.
surface mount caps come in standard sizes, yes. 0402, 0603, 0201, 01005, etc.
Scorpio had 2560 SPs with 2 CUs disabled and while the Xbox Scarlett APU is clearly bigger, it is not possible to speculate based just on the die size because of the different process node involved (7nm vs 16nm) but we can safely say that you are looking at at least 50% more power in the same die space. Since the Scorpio APU can output roughly 6 TFLOPs of power, we can guesstimate a range of around 9 TFLOPs for the Scarlett APU.
Huh that would be a twist lol gcn people feeling hopeful
Guessing these are 0612 given the aspect ratio.
That's based on his estimate of what the GPU size would be, without taking into account clockspeed increases.
I was being sarcastic :P
I thought it was decent but there's no denying it was heavily rushed to wrap up the trilogy and some plot points in previous films were either dropped or retconned
Everything is software in computers (besides silicon, and even then it's debatable for a large part of it).
Providing the old version of the software abstraction layer is not "software" in the sense that's it's basically the same piece of software that was present years ago, just still made available, sometimes with a small trick to be able to access it behind the latest version. There's no specific development needed to make it work on the latest hardware. I mean of course there are newer drivers for new hardware like GPU, but they're not considered "backwards compatibility software", just adjusted software to interface with the new hardware.
And yeah the whole point of PCs especially Windows PC is backwards compatibility. And of course sometimes you need special abstraction layers like WoW64 (to run old 32bit apps) but if you wanna run a 5 year old software on the latest i9 which didn't exist at the time, no special software layer is needed. That's the whole point. It just runs as the CPU presents the same instruction sets (and more).
401mm2 from Usman. Personally I think too high.
That's based on his estimate of what the GPU size would be, without taking into account clockspeed increases.
9tf would be hugely disappointing if true. Wonder where that number is coming from?
At the same clocks as Xbox One X
That article is claiming Xbox will be on stage at AMD conference almost as if it is confirmed. Huh?
That article is claiming Xbox will be on stage at AMD conference almost as if it is confirmed. Huh?
That's based on his estimate of what the GPU size would be, without taking into account clockspeed increases.
Phil has confirmed he's in Redmond and Tom stated in this thread he doubts there will be any any news.Wasn't Spencer on stage during last year's conference? Wouldn't be shocked if he showed up again to talk about series x and PC gaming in general
9tf RDNA is roughly double 6tf GCU, so it makes sense if the SeX is ~9tf. Would also make sense if the PS5 was ~10tf RDNA as well. Very close in power like everyone reputable has stated.
400mm2 pretty much confirms a 12 TF monster, IMO. 56/60 CU