Still, it will at most be Navi 14 which is small Navi. Likely less then 20CUa. And of course you wont be seing 8core APU in laptop paired with 16GB of RAM.Renoir was originally listed as Vega but more recently it is being shown as Navi.
Still, it will at most be Navi 14 which is small Navi. Likely less then 20CUa. And of course you wont be seing 8core APU in laptop paired with 16GB of RAM.Renoir was originally listed as Vega but more recently it is being shown as Navi.
Yeah there still is contradictory info about the GPU uArch of Renoir. Two things off the bat that for me make it highly unlikely Gonzalo/Flute is a AMD APU are 1, I believe they are 4c8t and 2, surely won't have 40CUs Gonzalo/Flute seem to have? Pretty sure there are some other things that point away from it too.
Raven Ridge used D or M indicators in it's OPN. G is for something more specific. Like a Subor or PS5.AMD Renoir APUs - 7nm/Navi/Zen2
AMD bin their chips, so Renoir performance probably ranges from poor to good/great. I very much doubt it's a single product configuration unlike Intel chips.
At the top end of Renoir's performance curve I could see it reaching the levels of Gonzalo/Flute - which I don't believe are next-gen console chips but ymmv.
I just simply believe it highly unlikely AMD go from the top desktop APU of the 3400G with 4c8t Zen+ and 11 CU Vega that launched 3 weeks ago to a 8c16t Zen 2 and 40CU Navi less than 6 months later.
Even with the cut down cache 8 cores seems top end for APU.. and why was there a single APU score, when we should expected multiple Renoir skus .. no this seems like a custom sample, not part of a product range
Depending of who you ask: Gonzolo believers or heretics
This alienware guy was recruted when?
Raven Ridge used D or M indicators in it's OPN. G is for something more specific. Like a Subor or PS5.
I'm thinking that they will be on par with one another.So the consensus so far in terms of power is PS5 > Xbox Next? Cerny is a genius if so. I won't be getting Xbox next gen if so.
Every today GPU has FP16... hell even older GPUs from 20 years (Nvidia FX even run FP16 twice faster than FP32 like Vega).Aaaaand you are wrong.
Polaris does not have FP16. Vega has FP16.
Since PS4 Pro has FP16, it cant be Polaris based.
Except it is.
It's full of obvious nonsense even for back then (like the 32bit CPU, no-HDD, and 'opengl gpu') mixed with some bits he must have overheard from a real source (like the codenames, 8 cores, and even the early 4 core part (which was not a Jaguar afaik)).Missed up a bit? The entirety of the leak is wrong Lol.
Which seems to be a running theme for early leaks - people get one or two pieces of random info and pad it out with their own fantasies.
This is what makes me think the 5700 in the consoles may be able to support RT.Every today GPU has FP16... hell even older GPUs from 20 years (Nvidia FX even run FP16 twice faster than FP32 like Vega).
What Vega has and Polaris not is RPM... Rapid Package Math... in simple terms Vega runs FP16 twice faster than Polaris.
PS4 Pro chip is Polaris based with some features from Vega but the main parts are Polaris.
So someone from Era cought one super interesting tidbit.
Flute and Gonzalo are both Shakespear characters lol that cannot be coincidence!
Ok so I checked for Ariel! Its Shakespear character as well holy shit
Ariel is a spirit who appears in William Shakespeare's play The Tempest
https://en.m.wikipedia.org/wiki/Ariel_(The_Tempest)
Francis Flute is a character in William Shakespeare's A Midsummer Night's Dream
https://en.m.wikipedia.org/wiki/Francis_Flute
Gonzalo (/ˈɡɒnzəloʊ/ GON-zə-loh) is a fictional character in William Shakespeare's The Tempest.
https://en.m.wikipedia.org/wiki/Gonzalo_(The_Tempest)
OMFG
RIP momma. The console wars miss you.Gonzalo believers have less base than my dead momma, is ridiculous, no thx :c
The heretics are just working with theoric facts, but nothing directly from employees, not from Sony, amd, or whatever.
Sigh ;;
As a person who studied practically all of Shakespeare's plays and sonnets, I'm personally embarrassed that I didn't decipher this myself.
Any guesses in whether the project will be a tragedy, comedy, or historical piece of work?
Any guesses in whether the project will be a tragedy, comedy, or historical piece of work?
mmmm Project Quantum revived?![]()
Alienware co-founder leaves Dell to become AMD’s chief gaming architect
Frank Azor is taking up a new challenge after 21 years at Alienware.www.theverge.com
![]()
Exclusive: Frank Azor Joins AMD As Chief Gaming Officer
Alienware cofounder Frank Azor has left Dell to join AMD as Chief Gaming Officer. Looks like AMD took a leaf out of Intel's book and went out on a talent hunt of its own, managing to grab Alienware Co-Founder, Vice President and General Manager, Frank Azor.wccftech.com
Renoir is laptop APU with low/mid Vega chip.
R600 is right, commercial AMD APU won't have nowhere the grunt of Navi 10, maaaaaybe Navi 14 and even that its probably too muchRenoir was originally listed as Vega but more recently it is being shown as Navi.
Unlike RPM, RT its not simple addition, it requires the GPU pipeline be designed around it.This is what makes me think the 5700 in the consoles may be able to support RT.
Curious about that too.mmmm Project Quantum revived?
Let me inject some data here. While pricing per wafer is hard to come by, and likely depends a bit on the specific chip (number of exposures) and of course volumes, current pricing seems to be just over $5000/wafer for TSMC 16/12nm and "just under $10000/wafer" (I'd ballpark it at $8/9000 or so for a 10000 wafer order) for TSMC 7nm. Again, at this point in time, as cost per wafer typically drops at a somewhat exponential rate after introduction. Also if you come in with a large order (say 300000 wafers) and stretched over time, that should affect pricing downwards as well.
Then you need to estimate the number of dies per wafer, and for a ballpark estimation you can use a calculator such as this.
Yeilds will tend to improve over time, and you would also design a big console chip with redundancy, for instance only using 7 out of 8 cores, or 44 out of 48 CUs or some such to mitigate the impact of defects.
So what's today's TF count?
Either 15 GC tapped together, or 2 Amstrad + 1 Pica Co-Processor.
More or Less.
Haha. That's all care about too. All the tech spec talk is way over my head.So what's today's TF count?
I just want TF and bandwidthHaha. That's all care about too. All the tech spec talk is way over my head.
Give me TF. Or give me polygons per second!
Funny how on ree they are pushing the narrative that 1700x performance is somehow good news
Interesting post on by3d
I used his estimates to calculate the costs of a ~390mm2 die
[..]
$9000/wafer, cost per die = 9000 / 115 = $78.3
Granted this is the cost straight from the fab without accounting for R&D, packaging, AMDs cut etc, lets add $50 and round it up to $130 per APU.
Also worth pointing out consoles manufacturers will place orders north of 100k wafers for the first year alone, so better better pricing per wafer
I'm looking for a time to triangle Cerny talk. All I want for my birthday.I just want TF and bandwidth
In their defense, 1700X is now much better CPU then at release, compared to competiton.Funny how on ree they are pushing the narrative that 1700x performance is somehow good news![]()
9TF max on 512-576GB/s of BW.I just want TF and bandwidth
I just went with by3d poster config, im guessing is the min size for detecting defects?Was defect density a guess, or are there up to date figures ?
Makes no sense from a performance or economic perspectivePS5 with dedicated ray-tracing chip/chiplet? I'm surely daydreaming, but here's my crazy theory: if that rumor turns out to be true
a glimpse of EARLY next gen games, totally forgot about this (11 months old)
At 18:52
PS5 with dedicated ray-tracing chip/chiplet? I'm surely daydreaming, but here's my crazy theory: if that rumor turns out to be true then maybe Microsoft and Sony won't focus on rasterization performance that much and instead they will spend their budget on raytracing. That would explain Gozalo being only equivalent to a 5700/XT. Imagine having a big dedicated chip for fullscene raytracing. With consoles being designed to last about 7 years, focusing on RT kinda makes sense. I bet many games in some years will switch from using rasterization/Hybrid RT.
Another point brought up in the video is that raytracing requires a lot of memory. Maybe next-gen consoles will have a lot of memory after all (or maybe that's why they'll probably be using the SSD as virtual RAM).
What do you guys think?
i already suggested it to(or maybe that's why they'll probably be using the SSD as virtual RAM).
What do you guys think?
i think hes right on this oneIt could but that would work against sony pr of no loading times
For 900 euros? Pshhh 26tf.There you go bois. Your 13TF machine with dedicated RT chiplet.
ARE YOU NOT ENTERTAINED?????
Already up for reservation for €900 euro's.
![]()
There you go bois. Your 13TF machine with dedicated RT chiplet.
ARE YOU NOT ENTERTAINED?????
There you go bois. Your 13TF machine with dedicated RT chiplet.
ARE YOU NOT ENTERTAINED?????
How come chiplets are advantage, when it's about CPUs, but suddenly not the right thing, when it is about APUs?Makes no sense from a performance or economic perspective
Its a shitty theory, sorry to be so blunt but it has no merit
Yes, something FORCED, when we hit 4Gb wall and were left with only ugly memory addressing solutions.But we're talking about x86-64, which is 16 years old.![]()