It’s quite easy to guess the SX speed based on the talk that was going around. Once you know 12tf, and 12.1 was assumed from like a year ago, there are only a couple of logical conclusions to make, and the most logical of those was 52 and 56 (and 52 based on past efforts and double the one x chatter) From there it’s backwards math to get your speed. It’s quite easy to do, to be fair. He wasn’t even the first person to say that number. And couples with all the many things he did get wrong... I don’t buy it. Now ignoring what I myself have heard and just going on as a person, that would count massively against the ps5 spec, too. Why would somebody know both, when 99% of everything else you have said turned out wrong? (Whats even more funny is that the rumour of 10.64 for ps5 was around November or so last year, and 10.64 x 1.25 is... 13.3, bang on. Plus when he gave his info, he gave 32kb l1 cache which is, you guessed it, GCN. How would he have 32kb cache and yet rdna figures? Think about it. So it’s likely he simply took that original ancient 10.64 rumour and just rounded it up to rdna, and went with it, without any actual knowledge as to what he was guessing. The numbers add up too well, but they added up at a time that rdna2 wasn’t a thing for either console. He later added rdna2 as a guess. So yeah, I don’t buy it. It all smells bad. To me, it seems like he caught wind of an old rumour and retrofitted it for his needs knowing a 1.25 gain for rdna 2, but didn’t actually know anything recent. That’s why most realistic examples are so different. It doesn’t make sense. What’s happened is that he saw what others were posting (again) and posted it, changing it to be 1.25, without understanding if the original 10.64 rumour he read was accurate It would need a complete system rewrite.)
But that’s a discussion for a later date.