I see no credible source lockhart ever existed.
Brad Sams is extremely credible. He was the first to give out the Project Scarlett code name.
I see no credible source lockhart ever existed.
Would 2080 with clock of around ~1500-1600MHZ be disappointing for consoles in 2020? I mean, it would only be ~9TF.It doesn't matter how we spin this, GCN flops < RDNA flops and all, but 36 CUs for not even 9Tflops would be pretty disappointing in 2020. I think 36CUs could have being a 2019 plan, but that plan would have changed.
Brad Sams is extremely credible. He was the first to give out the Project Scarlett code name.
It doesn't matter how we spin this, GCN flops < RDNA flops and all, but 36 CUs for not even 9Tflops would be pretty disappointing in 2020. I think 36CUs could have being a 2019 plan, but that plan would have changed.
I'd be surprised if AMD didn't have larger variants at some point or even an HBM variant for workstations.Any chance of >2 Navi Shader Engines?
Lockhart has been dead for greater than 6 months according to a few insiders. This is around the time final specs get nailed which is why you hear the whispers of who is more powerful
Where are people deriving the size of Gonzalo to be 310 mm?
There was only PCB leak in May describing entire motheboard for alleged PS5 dev kit, with die size of ~22.4 x 14.1, so probably 316 - 320mm².
Conceptually, but not physically?
This is the only source for that Gonzalo size. Anyone who can look up parts numbers could have created it. It conflicts pretty directly with Sony using the SSD tech from that patent, or the idea that the PS5 uses HBM+DDR4. So basically, people are taking that size estimate far more seriously than is justified.
To call Brad Sams not credible is a bit much. We all know things change in 12 months. My feeling is that the codename & leak was meant to gauge public and/or developer interest. It wasn't well liked or used how they expected, so they changed their plans. Doesn't mean the original source is not credible.
Tommy McClain
We all know things change in 12 months.
Being used as someone's flunky for controlled leaks is no better. Then having a glory complex that somehow you're in the inner-circle, just to have the rug pulled from under you... makes you a foolish flunky.
Or it's the Google Stadia effect, when they released specs at 10.7TF, MS changed their plans accordingly?
What leak is that HBM2 + DDR4? Not every leak is the same.This is the only source for that Gonzalo size. Anyone who can look up parts numbers could have created it. It conflicts pretty directly with Sony using the SSD tech from that patent, or the idea that the PS5 uses HBM+DDR4. So basically, people are taking that size estimate far more seriously than is justified.
I'm not even talking about HBM in regards to a specific rumor, just as a technical possibility. Buying into that measurement basically precludes belief in anything but GDDR6, and a standard pcie nvme ssd.
https://news.samsung.com/global/sam...6-gigabit-gddr6-for-advanced-graphics-systemsThe HBM2 with INFO_MS rumor is contradictory with Digitimes's info that PS5 will be packaged at ASE. INFO_MS is a TSMC proprietary packaging solution and it's highly unlikely to be used for non-mobile devices.
Integrated fanout as far as I know is used for mobile devices only. I am 99% confident that PS5 is a flip chip.
The only inconsistency in the OQA PCB leak is the 18gbps 16gb GDDR6 chips that are not even listed on Samsung's site.
https://news.samsung.com/global/sam...6-gigabit-gddr6-for-advanced-graphics-systems
Though Samsung mentions it in their newsfeed from January.
I think this only further strenghtens the rumor, as this is something not many would be aware of.
I am relatively sure if you are big console manufacturer you dont procure on their product pageI am aware of the newsfeed. Just find it suspicious that 16gbps is in sample, but 18gbps is not even listed. :/