PS4 Pro Speculation (PS4K NEO Kaio-Ken-Kutaragi-Kaz Neo-san)
- Thread starter mpg1
- Start date
-
- Tags
- sony sony roach motel
- Status
Metal_Spirit
Regular
Dear Sir. We are talking assync computing here. None of the games on the benchmarks you sent are known to use it.
And my words were "on some games".
You are in the wrong context!
Inuhanyou
Veteran
That is the general jist i've heard. While Scorpio would be using Vega
Every console GPU is underclocked below their PC variants because of heat or more favourable binning. We knew for months from the Neo leak that its GPU will be clocked at 911MHz. RX 480 is clocked 300mhz higher on PC.
Lowering clocks will make Neo APU run cooler and pull less power, it is a sensible choice for a console that will have very small casing.
I'll be very happy with the Neo if that turns out to be the case, 970 equivalent perf in a closed console? That'd be great, especially for 1080p 
It's definitely possible in terms of power consumption too, the launch Ps4 was like this
I'd imagine an under-clocked 480 and a zen CPU would be around the same metrics, if not less than 150W. The interesting thing is that the 7850 (a bit worse than Ps4 GPU) cost around $160 around the release of the Ps4 so the 480 in the Neo is most probably doable (cost wise).
It's definitely possible in terms of power consumption too, the launch Ps4 was like this
I'd imagine an under-clocked 480 and a zen CPU would be around the same metrics, if not less than 150W. The interesting thing is that the 7850 (a bit worse than Ps4 GPU) cost around $160 around the release of the Ps4 so the 480 in the Neo is most probably doable (cost wise).
Well they seem to do so on PCs which have more regular and smaller updates?
Like what? Add-ons generally fail and that's because you split the userbase, Neo/Scorpio are aiming not to do this.
If small form factor is their main concern thus the down clocking, would people really care if Neo is slightly bigger than OG PS4, say Xbone sized console? It's clear you've got enough CUs to reach 5.5tf, even upclock it to ~5tf would be welcoming. Just wondering if this is purely a design choice for aesthetics or something else.
Bigger design means more plastic, high clock speed probably means more voltage which means more heat which means more money spent on a cooling solution plus a more capable PSU to handle the higher TDP. It might be a combination of both aesthetics of the console and additional expenses (however significant or insignificant they might be).
If, Sony is aiming Neo at the higher-end segment of the market (diehard gamers), they shouldn't be tampering with lowering clock speeds, and worried about TDP. Build the system (NEO) around what the GPU is fully capable of ...even if it's slightly louder and more power hungry than PS4. Console gamers really need to man up in these areas...
Totally. I feel a wave of power washing over me when I hear my PS4 actually doing some work and kicking up the fan! You better work, bitch!
I consider the idea of foward/back copatible progressively upgrading consoles great and totally plausible but there is more to mere flops and architecture to executing it right.
My stance is, the way this rumors are painting the picture, devs had little head up for this. PC games are developed from the ground up with hardware compatibility in mind, they run at a higher level, through windows, dx, and the video card's drivers, and even then fuckups are not uncommon.
I bet console engines and projects lack a lot of the infrastructure required to be able to just easily improve gfx quality substantially. Sure things like resolution and draw distance are just a variable they can modify right? Until they do and a 100 bugs pop out of nowhere. Oh it seems this gameplay designer coded this script on this level assuming this objext would be loaded into view at this distance. Oh, this system bases its calculations on specific pixel numbers over here, but later it would use the relative screen coords. Gotta rewrite that... game dev is messy, and deadlines and crunch leaves a dirty trail of shortcuts and hacks all over the place. Even multiplats released on consoles, run in a branch of the engine that rids itself of a lot of the expensive infrastructure of the pc version. That has to be put back in now. This are simple changes from an enginerimg pov, certainly, but on a large AAA game, full of features and content, and a large codevase, they take weeks, if not months. Most devs will half ass it initially. And MS will be up there trying to sell something as big as the bone and louder, costing 600 bucks, that is suposedly a beast, and all they have to show is the same games of the regular bone at slightly higher res and better fps.
The avarge gamer will see no difderence. Digital foundry is not a console seller.
Lets even ignore the technicalities, and suppose developers can just mark a few checkboxes in their engine as if it were a pc gane and out is the Kaio Ken and One Two versions.
What would the graphical enhancements be? Longer draw distance, better SSAO, aniso, texture streaming, higher-res post processing... basically the difference between a game in ps4 and a pc on Ultra everything. I can sure spot the difderence (with some effort actually, this day and age) but do you think the avarage joe gamer sees any difderence at all.
Lets get our collective heads out of our own niche beyond3d-member digitalfoundry-reading siggraph-presentarion-watching graphics-enthusiast asses and think of the large mass of consumers. You can grab a FAT majority of them, show a game in ps4 and pc ultra 4k side by side, and they will still tell you they look the same thing. They really don't care. They just wanna play the game.
So, do I think there is no place for a premium machine? No, I think there is. Even if a consumer is a layman, get a salesman to tell them X is MOAR BETAR than Z, those with enough money will buy into it. But there is a sweet-spot on how much more expensive that premium version can be before it becomes too niche. And the more niche, the less devs will care, thus their improvementes staying less substantial, making it even more niche in a vicious cycle.
The only reason I saw ps4 kaio ken as possibly working commerecially, was if it downplayed the improved graphics, and really just sold it as a ps4 that can play 4k movies, has improved media features x y and z, and oh, by the way, your ps4 games will play at 4K* too (*:1080+ upscaled with 4k hud). I know people here had much more ambitious speculations for the thing, but I was personally very sceptical of the marketability of that kind of thing, and of the willingness of devs to put the effort in to achieve them. As for Xbone two, I think 5-6Tf is bound to go mostly to waste.
Last edited:
Bagel seed
Veteran
RancidLunchmeat
Veteran
You really believe Sony is still going to launch the PS4K aka PS4.5 Neo when MS is releasing a real next-gen console a few months later?
Fall 2017 vs fall 2016 isn't few months away and i don't think Sony really care at this point, the Playstation brand is stronger than ever in the two competing markets where Xbox means something, the UK and US, it's a massacre everywhere else. Also, the pricing will be different for a 6tf console vs a 4tf console if both decide to go this route, they won't be competing against each other. And price is crucial (as the Wii/360 have taught us last gen, or the Ps2 before them or ...).
Last edited:
upnorthsox
Veteran
You mean do I really believe that MS is going to release a Vega/Zen/HMB2 500mm wonderkind? No, if the rumor is real then it'll be basically the same chip AMD has offered to Sony with maybe some modestly higher clocks. So at 5tflops, is Sony about to cancel plans? Esp. when they could mildly boost theirs up to 4.6Tflops (1000mhz) too? And this 4 yrs into a 6 yr gen with a 1 yr lead? Not a chance.
Yes. Sony wants the best possible start for PSVR. I'm almost sure the breakout box components are either integrated into the Neo design or the GPU is beefy enough on handling all PSVR duties. Either way, Sony is going to capture this year's holiday sales, and contend with what Microsoft has in late 2017.
- Status