Brand new account. Every post in this thread. Who's alt are you?
lol he sounds like me
Brand new account. Every post in this thread. Who's alt are you?
Andrew Reiner is an editor at Game Informer, and tweeted that prior to E3 2019, developers were telling him that the PS5 was more powerful than Project Scarlett (as it was known at the time):Never heard of em, but yeah that sucks. What did they leak if you don't mind me asking?
1. I ask because there are some things that for me do not add up. Why would they be testing for things like backwards compatibility on dev kits? What does that achieve when they have another chip?
2. What developers get early on is target specs. They then usually get PC parts to start development as development kits are prepared an iterated upon. Nothing strange. Now the question is, if developers got dev kits in and around 13TF RDNA after GDC, why is this not being tested?
Wow, this makes a lot of sense.Allow me to take stab at a scenario based on the theory I've mentioned earlier...
1. Sometime in 2018 or so AMD creates a devkit for Sony that's in a big silver PC case. This devkit is running on early samples of the 5700 GPU and Zen 2 CPU, and is using their in-development SmartShift tech to run as an APU for a performance boost and for compatibility for the devs. Up until this point that's the most powerful hardware AMD has available running Navi and Zen2 and lets devs get an early start on hardware that's closer to the final hardware than anything else available.
2. In early 2019 new silicon is available and the next devkit in the V-shaped form factor becomes available. This devkit goes out after GDC and might be an actual APU. This is why people at GDC 2019 were reporting 9.2 TF, because up until that point they only had access to the low speed devkit. It wasn't until after GDC 2019 when the V-shaped devkits went out did we start hearing about figures upwards of 13 TF from Benji and later Tom Warren when he was reporting from E3.
3. In April 2019 Sony talks to Wired and is using the earlier devkit in the PC tower in their presentation. This unit was specifically referred to as a low speed devkit.
4. In June or so this AMD engineer logs a test done on a devkit, not long after Cerny showed they were still using the older low speed devkits. It's entirely possible this test was done on the PC tower devkit. Why? Well I'd imagine AMD has to test each devkit before sending them out. The GPU could have been clocked at 2GHz as a stress test as well instead of always running at 2GHz. This is why the Github hardware seems to closely match a 5700 GPU and why there's no hardware raytracing to be found. As for why would they still be sending out the low speed devkit when the new V shaped unit just became available? I can guess a few possibilities. Not every studio is equal and some of them might not get the hardware that's closer to the final product that early for any number of security reasons. Also not every employee in a studio is equal so only certain employees might have been using the new V shaped devkit while others could still do their job just fine on the low speed devkit. Or maybe at the time they still couldn't produce enough V shaped devkits so they had to still keep pumping out PC tower devkits to keep up with the demand.
So there you go, what I think is a plausible explanation for what the Github devkit was. It explains why its a fair bit less powerful than insiders have indicated. It explains the timing of why we were hearing 9.2 TF before and at GDC and upwards of 13 TF after GDC. It explains why the Github unit seems to basically be a Zen 2 CPU and a 5700 GPU slapped together and why there's no raytracing in the tests. It also explains why Matt indicated the Github data is no longer relevant.
If anyone wants to poke holes in the theory then by all means. It's by no means perfect, but I think it's a decent enough guess and pretty plausible.
Where are all these fanboys leaking from? Both sides are awful in here. Everyone needs to chill on which piece of plastic they prefer.
My problem with gcn vs rdna is that devs would not say they are same if one is 12 rdna tf and one is 12 gcn tf . (If I understood your second point correctly.)
True. But we could just add them together and average it out.Actually I think we need 2 specs, one for GPU computational performance and one for ray tracing performance. I would like them disjunctive otherwise you muddling the waters.
1. I ask because there are some things that for me do not add up. Why would they be testing for things like backwards compatibility on dev kits? What does that achieve when they have another chip?
2. What developers get early on is target specs. They then usually get PC parts to start development as development kits are prepared an iterated upon. Nothing strange. Now the question is, if developers got dev kits in and around 13TF RDNA after GDC, why is this not being tested?
I'm still not convinced. A cv from XING or LinkedIn with a bit of fantasy and then a public repo of all things? That's the equivalent to some self printed notes to me tbh. We have people creating short animations of a fake PS5 UI with a console and some fake documents or 3D printing a fake Nintendo Switch.I don't recall the full details, but IIRC it was found with the full resume of the AMD engineer who logged the data.
if developers got dev kits in and around 13TF RDNA after GDC, why is this not being tested?
I'm still not convinced. A cv from XING or LinkedIn with a bit of fantasy and then a public repo of all things? That's the equivalent to some self printed notes to me tbh. We have people creating short animations of a fake PS5 UI with a console and some fake documents or 3D printing a fake Nintendo Switch.
No, I'm only going to believe what established journalists like Schreier say or even better, I'll wait for the official announcement.
Who's to say that it wasn't tested. Those tests just haven't leaked.
I don't know about you guys, but I for one can't wait to see PS5 be more powerful than XSX, and seeing Xbox execs do damage control at first but then go back and improve the whole platform in other ways the upcoming generation, just as this generation.
I mean when has boasting about sitting on the power throne ended with something positive for the end user?
Well we do know that ps5 dev kit around gdc 2019 was 9.2 tf (before the v shape dev kit)1. No idea. I don't recall much about the BC portion of the Github leaks being discussed and I never saw them myself.
2. Well in point 4 of my guess I noted it could have been the old PC tower low speed devkit that was being tested because AMD would have to test each unit before sending them out, and running them at 2GHz could've been a stress test. There're any number of reasons why they could've still been sending out low speed units when the V shaped devkit just became available.
I know there are alternate theories out there, but I like this one because it fits so nicely with the timing of all the events before and after GDC 2019. I think the V shaped devkit did clock in at about 13 TF which explains all the rumors we heard from insiders over the past year. The XSX is likely in that same ballpark which is why so many people are giving a power difference of 10-15%.
Naah. Microsoft can't afford to have the weakest box against at the start of a generation. So even if both PS5 and SeX are currently very similar, Microsoft will wait for the PS5 reveal before bumping the clock up. Not that it will matter at the end, next gen will be about more than just TF.I don't know about you guys, but I for one can't wait to see PS5 be more powerful than XSX, and seeing Xbox execs do damage control at first but then go back and improve the whole platform in other ways the upcoming generation, just as this generation.
I mean when has boasting about sitting on the power throne ended with something positive for the end user?
Well we do know that ps5 dev kit around gdc 2019 was 9.2 tf (before the v shape dev kit)
Well we do know that ps5 dev kit around gdc 2019 was 9.2 tf (before the v shape dev kit)
Yea the v shape went out to devs one month after gdc 2019 around April 20.and apparently there are two v shape dev kit prototypes . One for first party and one for 3rd party( different specs to prevent leaks )Yup, that had to be the low speed devkit in the silver PC tower. As I mentioned earlier, if the V shaped unit went out just after GDC then that means this silver PC tower must've been created quite a while before that. Definitely in 2018, maybe as far back as 2017? What sort of hardware would they have had back in 2018 to shove into a devkit? I'm seriously doubting they would already have an APU more than 2 years away from launch, but they almost certainly would've had early 5700 GPUs and Zen 2 CPUs since those were launching in mid 2019.
Now I didn't read the Github leak and all the info I have on it comes from the others in this thread. To anyone who did read it, is there really any reason why the device being tested there couldn't have been a 5700 GPU OC'd possibly for a stress test and a Zen 2 CPU, then use SmartShift to make the system think it's an APU?
Yea the v shape went out to devs one month after gdc 2019 around April 20.and apparently there are two v shape dev kit prototypes . One for first party and one for 3rd party
it's a rumor thought . I ll post it .Wait, 2 V shaped devkits? I don't recall hearing about those. Source?
it's a rumor thought . I ll post it .
Prototype 1 : prospero (3rd party)
Prototype 2: Verona (1st party)
4 more days. If we don't get an announcement then then #teamfeb needs to hold their heads in shame!
Spoiler alert: It ain't happening in February.
4 more days. If we don't get an announcement then then #teamfeb needs to hold their heads in shame!
Spoiler alert: It ain't happening in February.
4 more days. If we don't get an announcement then then #teamfeb needs to hold their heads in shame!
Spoiler alert: It ain't happening in February.
Interesting to see whether those who don't believe ps5 specs may have evolved are willing to countenance that Lockhart might have. If anything the latter would make more sense if ps5 was on a level with Xsx
Very well thought out. Nice post Viv.Allow me to take stab at a scenario based on the theory I've mentioned earlier...
1. Sometime in 2018 or so AMD creates a devkit for Sony that's in a big silver PC case. This devkit is running on early samples of the 5700 GPU and Zen 2 CPU, and is using their in-development SmartShift tech to run as an APU for a performance boost and for compatibility for the devs. Up until this point that's the most powerful hardware AMD has available running Navi and Zen2 and lets devs get an early start on hardware that's closer to the final hardware than anything else available.
2. In early 2019 new silicon is available and the next devkit in the V-shaped form factor becomes available. This devkit goes out after GDC and might be an actual APU. This is why people at GDC 2019 were reporting 9.2 TF, because up until that point they only had access to the low speed devkit. It wasn't until after GDC 2019 when the V-shaped devkits went out did we start hearing about figures upwards of 13 TF from Benji and later Tom Warren when he was reporting from E3.
3. In April 2019 Sony talks to Wired and is using the earlier devkit in the PC tower in their presentation. This unit was specifically referred to as a low speed devkit.
4. In June or so this AMD engineer logs a test done on a devkit, not long after Cerny showed they were still using the older low speed devkits. It's entirely possible this test was done on the PC tower devkit. Why? Well I'd imagine AMD has to test each devkit before sending them out. The GPU could have been clocked at 2GHz as a stress test as well instead of always running at 2GHz. This is why the Github hardware seems to closely match a 5700 GPU and why there's no hardware raytracing to be found. As for why would they still be sending out the low speed devkit when the new V shaped unit just became available? I can guess a few possibilities. Not every studio is equal and some of them might not get the hardware that's closer to the final product that early for any number of security reasons. Also not every employee in a studio is equal so only certain employees might have been using the new V shaped devkit while others could still do their job just fine on the low speed devkit. Or maybe at the time they still couldn't produce enough V shaped devkits so they had to still keep pumping out PC tower devkits to keep up with the demand.
So there you go, what I think is a plausible explanation for what the Github devkit was. It explains why its a fair bit less powerful than insiders have indicated. It explains the timing of why we were hearing 9.2 TF before and at GDC and upwards of 13 TF after GDC. It explains why the Github unit seems to basically be a Zen 2 CPU and a 5700 GPU slapped together and why there's no raytracing in the tests. It also explains why Matt indicated the Github data is no longer relevant.
If anyone wants to poke holes in the theory then by all means. It's by no means perfect, but I think it's a decent enough guess and pretty plausible.
Exactly the only way ms would bump the spec of Lockhart is that they are sure ps5 and xsx are close in spec and price and now they feel they can price Lockhart 399 and leave 499 for ps5 and xsxInteresting to see whether those who don't believe ps5 specs may have evolved are willing to countenance that Lockhart might have. If anything the latter would make more sense if ps5 was on a level with Xsx
Have they ever done a 2 weeks notice for an event before?4 more days for what? Sony could go through to 14 Feb or they wanted to give 2 weeks notice for an event last week of Feb.
4 days is only if they are literally copying 2013 dates
But to be honest, they ARE.A month notice is common as people need to book flights and plan accordingly. 2 weeks sounds rather arrogant as if everybody just sits there and is waiting only for this invitation.
You didn't answer his question . VN1XA month notice is common as people need to book flights and plan accordingly. 2 weeks sounds rather arrogant as if everybody just sits there and is waiting only for this invitation.
They didn't cancel Destination PlayStation, looks like that was a bad rumor.It more than likely end of February or before GDC in March, they cancelled their appearance at Destination PlayStation which is usually held in February and everything will leak if they leave it after GDC.
Price will not be revealed till June / July you can bet on it imoIMO we'll get an article or a state of play, with price and some game montage sometime in Feb and then a full blowout in May/June.
Yeah before the 31st of Feb which is when they announced the event date for the PS4 reveal.
I wanted to write "specs" not price, damn. I think they'll give us specs in an article and then some kind of video to show it off.Price will not be revealed till June / July you can bet on it imo
31 of Jan . Yeah . For feb 20 eventYeah before the 31st of Feb which is when they announced the event date for the PS4 reveal.
They didn't cancel Destination PlayStation, looks like that was a bad rumor.
I can't really see what the purpose of putting an NDA on an invitation would be.Perhaps the press have already receive the invitation and sony put an NDA for reveal they are invite.
For reveal the event to the public when sony want to reveal.I can't really see what the purpose of putting an NDA on an invitation would be.