PS5 is doing all of that in a custom chip rather than on the CPU. It's part of how they can keep the whole flow so fast. He gave an example of CPU decompression performance would be needed to feed the SSD and it was fairly prohibitive basically.
It's a variable set... by definition, it is not doing anything 100% of the time and a "boost" can't be called a boost if it's the base speed and is always that speed.the presentation specifically stated they run boost mode 100% of the time.
Both of these consoles will be so blazingly fast, the difference in multiplat titles will be for one, unnoticable. What I gather from it, XBX is brute power, and PS5 does things in a far more intelligent fashion. both will have no problem driving incredible graphics to your TV's.
It don't work that way since certain parts of the GPU scale with speed.
That not to say PS5 is better but it's not as clear cut as that .
Ah, didn't see most people were skipping that Xbox is at 12.16TF locked
Sony didn't mention anything on that front. I would assume that clockspeed doesn't necessarily change with SMT, but that with the variable frequency the clock will veer away from 3.5GHz.That's what it sounds like.
We're we given any specific information on CPU clock speeds with SMT?
its always in boost mode, it just varies how much. its not a on and of thing.
The variable frequency is there so the PS5 doesn't draw full power and heats up while you are doing something that's not computationally intensive.if they ran 100% of the time there'd be no need for variable frequency
Microsoft been knew.Only unexpected thing was the pre-shade Microsoft threw by constantly focusing on ' fixed frequency ', they are truly forward compatible in their approach, lol.
Doom and gloom is unnecessary though, it's an adequately capable machine.
Pretty sure that's not the case. I think he said they tested the Top 100 games (purchased/rated, I'm not sure) and almost all worked. That makes it sound like 80%+ of all titles.
the presentation specifically stated they run boost mode 100% of the time.
In the BEST CASE scenario (for the PS5) the XSX has 18% better flop performance. And that's in the theoretical situation where the PS5 can sustain full boost clock indefinitely. Which it won't. The worst case scenario is probably very close to the GH leak, so around 9.2TF. In that scenario the XSX is around 31% ahead. The average case might be somewhere in-between. The XSX also has a faster CPU and faster memory (for its main pool of GPU memory).
All of this will show up in 3rd party games, no doubt about it. The SSD advantage of the PS5 can really only be used optimally in exclusives (which I'm sure will be amazing).
Looking like I need to get both of these to get the best of both worlds. Was hoping the PS5 would be all I needed.
Agreed there's some theoretical advantage, but it does rely on developers using the memory split in that specific way. You can't just say for that other 6 that it will only impact the CPU wholesale.yean, you cannot average the XSX GPU memory bw like this. The GPU data will reside on the optimal pool, while the slower pool is intented to be used for mostly CPU related stuff.
It is my understanding that DirectStorage on the XSX will not be using the CPU for decompression.
The variable frequency is there so the PS5 doesn't draw full power and heats up while you are doing something that's not computationally intensive.
its always in boost mode, it just varies how much. its not a on and of thing.
Because devs are telling him they are close.Why was Jason saying these consoles were "extremely close" when apparently he didn't actually know? Weird.
They tested the top 100 and "almost" all worked, whatever that means. The presumption being is that almost all games will work. What doesn't, if it's something they can address wasn't shared.Only 100 of the best PS4 games? Xbox has all Xbox one games day one? That's shit.
Only 100 of the best PS4 games? Xbox has all Xbox one games day one? That's shit.
I prefer XSE vs XSS, sounds less WW2ish.
It's just not loading, it's how the game engine is built around streaming data in dynamically and the density of data in the viewable space. Time will tell.Actually it's more than 20% in GPU power. Also about that 100% more SSD bandwidth will most likely result in difference of fraction of seconds, so not actually appreciable.
Though i agree 1600p vs 2100p may also not be appreciable.
So i wondered if i missed something, was there any mention of hardware specific for raytracing work?Raytracing is going to be huge this gen, if PS5 is crippled in that regard it's going to be even tougher news.
no I said it runs 100% in boost mode. the mode itself can throttle, but it tries to boost all the time, with temps being the influencing factor. they had to cap it to make the system quiet enough and cool enough. put the machine in a hot room, it will clock lower, put it in a cold room and it will boost to the cap.If it runs 100% of the time then it would be fixed. Come on people.
That's not what he said. As an example, he said of the top 100 games played on PS4, most run fine. That doesn't mean many others beyond that peak of the pyramid don't also run fine.
For the SSD drive, is the Xbox velocity architecture being considered?
No they are not.The differences are negligible, which makes the price discussion even more interesting.
no I said it runs 100% in boost mode. the mode itself can throttle, but it tries to boost all the time, with temps being the influencing factor. they had to cap it to make the system quiet enough and cool enough. put the machine in a hot room, it will clock lower, put it in a cold room and it will boost to the cap.
Did anyone watch this presentation? or are you all just looking at the spec sheet and inventing meaning from words.
The PS5 SSD speed is absolutely crazy. It's by a great length the biggest leap in configuration that one of these systems has over the other.