Middle Generation Console Upgrade Discussion [Scorpio, 4Pro]

Status
Not open for further replies.
why not just have minimum console gen on there? Person knows that their machine can play all games before their gen.
If you have Xbox one, and game says Xbox 2, you know it's not going to run on your machine.
All this may seem like it will be confusing to general public, but it will just be the new normal for Xbox, so they'll be fine.
The only problem is the initial change over to this new philosophy. And their console names, as calling it Xbox 2 will mean some people keep going on about it being a new gen, that noise may only matter in the immediate sort term though.
Also it only really matters on physical, and by time Scorpio follow up is due their not going to care that much about physical anyway.
People keep going on about is it worth having console just go pc, yet that's a lot more confusing, and all they say on front of case is pc.
 
why not just have minimum console gen on there? Person knows that their machine can play all games before their gen.
There will be a psychological aspect. Who wants to buy an XBox Two all new and shiny and yet play XBox One games on it? They'll want the new, improved XBox Two games. Smartest thing to do is probably have the same game and just a different stick on label which the stores can swap over...

Seriously though, there is marketing concern here. If MS can convince people Scorpio is a new generation, they'll get more sales. That'd require distancing themselves from the previous generation, so, for example, saying, "also backwards compatible with your XBox One library". However, if MS want to propagate the notion of XBox being a progressive platform then they don't want generations and even want a message that would disrupt Sony's attempt at a new generation. "New PlayStation - every 6 years. XBox - forever." Sell people the idea that their games will continue onto new boxes and get better as a result.

I think that's the better move, the harder initial sell and more challenging message, and the way to capitalise on MS's unique position. It's also the direction MS is more aligned with.
 
I agree about harder initial sell/explaining the message.

people buy pc games and all it says is pc on front, and they seem ok with that.
requirements on back would be extremely simple, minimum gen, recommend gen. None of the different system configurations which people have to guess how it performs on their set up.
they psychologically know they have premium machine.

becomes even simpler when digital.
I really only think it's an initial problem which could easily be offset by early adopters anyway
 
in 2022 - Xbox two 2017 _> xbox three 2021 xbox one support is now dropped with the last few games releasing.

in 2025 xbox two 2017 -> xbox three 2021 - > xbox four 2025 and in 2026 the last of the xbox two stuff comes out and its no logner supported with new games.
Whatever MS is doing with their *ecosystem*, Scorpio is still dying together with X1(just like Pro with PS4), or just a little bit later, i don't know how it's still not clear.
 
Why should it be clear? Spell out the argument.
I've tried to do it over the year:
Zen ultimately changes nothing, with PS5 Sony is going to have 7nm Zen+, that would be on another lvl performance-wise, that is along with HBM(low-cost), more (V)RAM, not to mention up to 14TF GPU and other stuff, providing actual generational difference over Scorpio(and PS4/Pro) just 2 years after its release(yeah, i believe 2019 is more likely now, as 2020 probably won't offer anything arch-wise(or such is the current outlook, anyway) and it would be worse timing).

Scorpio is caught between a rock and a hard place, there isn't much to do, besides "outside of the box" stuff, whatever that may be. It's not going to turn the industry upside down launching with a price of $499(12GB of GDDR5X(something they might rethink?) and a BD 4K drive), having no(?) top-end exclusives(Scorpio only).

They likely "lost" already by not releasing their "Pro" in 2016(MS probably didn't think Sony would, since releasing PS4 Pro this year is like releasing PS4 in 2012(node-wise)), that's the №1 thing that matters this *mid*-gen, *next*-gen logic does not apply.

Hardly anyone is going to abandon X1(and PS4 St./Pro, similarly powered devices, etc.), yet support Scorpio, when 7nm consoles hit the market, it's either gonna be 8th gen or 9th.
There'll be basically two tiers: Jaguar-based one, and Zen(PS5)-based one, with a rather clear divide, the only way to get to the second tier is to be it(duh).


For example, MS could've done an actual premium console CPU+GPU setup, Desktop 8-core Zen, 8GB DDR4, Vega10 8GB HBM2, $999, and even that product would(/could) be in a tough spot when a conventially priced PS5 releases.
 
The 4 core 8 thread gave it away. MS said "8 core" during Scorpio reveal which I assume to be physical not logical.
Ryzen sounds nice. Fitting the current console economy and power ratio, is tough. Maybe a derived product. Vega sounds more in the line of what to expect.

Be it Vega or something else, a feature I really want Scorpio to have is FreeSync. It would be the first console ever without tearing and stuttering.

Vega releases in May, so FreeSync sounds feasible.

http://wccftech.com/amd-vega-gpu-radeon-graphics-may-launch/

AMD-VEGA-10-specifications.jpg


AMD-VEGA-10-VEGA20-VEGA-11-NAVI-roadmap.jpg
 
I've tried to do it over the year:



There'll be basically two tiers: Jaguar-based one, and Zen(PS5)-based one, with a rather clear divide, the only way to get to the second tier is to be it(duh).


For example, MS could've done an actual premium console CPU+GPU setup, Desktop 8-core Zen, 8GB DDR4, Vega10 8GB HBM2, $999, and even that product would(/could) be in a tough spot when a conventially priced PS5 releases.

You're assuming that Microsoft won't release a new console at 7nm using relatively similar technology & price? If you think that Sony is planning to launch PS5 in 2019/2020 (with actual generational difference compared to Scorpio) then what makes you to think that MS won't release a 7nm version of Scorpio with much lower price tag (300-250$) alongside Scorpio 2 (which should be considered as PS5's direct competitor)?
 
You're assuming that Microsoft won't release a new console at 7nm using relatively similar technology & price? If you think that Sony is planning to launch PS5 in 2019/2020 (with actual generational difference compared to Scorpio) then what makes you to think that MS won't release a 7nm version of Scorpio with much lower price tag (300-250$) alongside Scorpio 2 (which should be considered as PS5's direct competitor)?
No, i was explaining why no stuff would be coming out in 2026 for Scorpio.

If MS wants to stay in the market, they would have to release their next box not long after PS5. I would expect 7nm shrinks of all current consoles.

Also, i find an idea of driver-less consoles sold at a loss entertaining, digital sales continue to increase(20%+ of the total, according to some reports), platform holders get more from a copy sold at their store, especially if it's their own game(100%), we could get $99 current-gen boxes and $399 next-gen boxes(with conventional model costing $499), the latter makes even more business sense, since the owner of it is likely spend more, much more, than the owner of a soon-to-be past-gen box.

But then again, why do anything, if the money go your way anyway ¯\_(ツ)_/¯
 
No, i was explaining why no stuff would be coming out in 2026 for Scorpio.
It's an argument but one very far from clear if you follow the discussion and the possibility of 1) PS5 not being that far ahead and 2) the value of a progressive platform meaning Scorpio is still an entry level for PS5 era games alongside a more costly, more full-featured XB2. Plenty of unknowns and legitimate alternatives rather than the shoe-in you believe in.
 
The problem with the "eternal console" concept is that it robs you of the "All New" message and leaves devs in an awkward spot when it comes to baselining a game. We even saw this to a limited extent this gen with the early years of PS4/XB1 being dominated by cross-gen titles that offered mildly better AA and texture res bumps. Now imagine that period never ended and that's the risk of the "eternal console", I accept this transition was particularly stark thanks to the 8 year life cycle for the former gen

The defined "Gen X + Gen X.5" model at least offers a firm cut off point to allow devs to stop designing for the limitations of the previous and just focus on what's available to hand, the ".5 gen" is relatively modest to account for during the life cycle and adds modest overhead for compliant support (you can go nuts but the base compliance seems modest). When we get to Gen Y we will of course have that same transition period but that will end and then devs can use the new and more radical (I hope) differences between Gen X and Gen Y more fully.

That's before we get into the perverse incentives a misfired gen leaves devs with in the eternal console, let's say we have V4 sell 50 million units over 5 years but V5 is only at 20 million units at 3 years and shipment volumes are soft, what do I focus on? I can presume that those 20 million users were V4 users so now I have a market split 3:2 in favour of the slower console, as a producer where should I allocate my dev resources? Do I put them into trying to really show off V5 and leave V4 users with a much reduced product (Shadows of Mordor on x360/PS3 leaps to mind) which given they are a plurality of the market seems like a risky move? Or do I decide instead to rework my gameplay and design to fit within V4 limitations and instead turn the V5 version into a de facto V4.5? Perhaps V5 winds up as a stunted gen but V6 enjoys vast success now V5 buyers have been left with a s/w gen that never really delivered for them and now they get downports for the next cycle.
 
  • Like
Reactions: JPT
It's an argument but one very far from clear if you follow the discussion and the possibility of 1) PS5 not being that far ahead and 2) the value of a progressive platform meaning Scorpio is still an entry level for PS5 era games alongside a more costly, more full-featured XB2. Plenty of unknowns and legitimate alternatives rather than the shoe-in you believe in.

Yep. Come 2019, I think it unlikely that a AMD's current flagship - Radeon 480 - will have fallen off the bottom of the minimum specs for multiplatform games. And Scorpio should be faster than that (it'll be newer and it certainly has a lot more bandwidth).

In terms of memory available to games, Scorpio should have ~9GB compared to the PS4Bone's 5GB, so there's lots of room to scale there in terms of minimum datasets. CPU is up in the air still, of course, but consoles do a remarkably good job of getting by with skimpy CPUs, even if the cost of that is 30 fps.

If Scorpio does die rapidly upon a 2019 introduction of PS5 it'll be because the market for software isn't there, not because multi platform games can't scale down to the system. 7nm should be around double the density of 14 nm, and less than double the performance at the same power envelope.
 
Ryzen sounds nice. Fitting the current console economy and power ratio, is tough. Maybe a derived product. Vega sounds more in the line of what to expect.

Be it Vega or something else, a feature I really want Scorpio to have is FreeSync. It would be the first console ever without tearing and stuttering.

Vega releases in May, so FreeSync sounds feasible.

http://wccftech.com/amd-vega-gpu-radeon-graphics-may-launch/

AMD-VEGA-10-specifications.jpg


AMD-VEGA-10-VEGA20-VEGA-11-NAVI-roadmap.jpg

Is that suggesting 1.46 ghz for Vega 10, or have I just done a maths fails?

[Edit] .... and so .... about .7 ~ .8 of that for a more power constrained console? So maybe 1.02 ~ 1.17 gHz for Scorpio?

[Edit 2] So mebbeh 44 active CUs and 4 redundant? 1.065 gHz for 6TF? Final clocks and TF to be determined at mass production of final silicon?
 
Last edited:
I've tried to do it over the year:



There'll be basically two tiers: Jaguar-based one, and Zen(PS5)-based one, with a rather clear divide, the only way to get to the second tier is to be it(duh).


For example, MS could've done an actual premium console CPU+GPU setup, Desktop 8-core Zen, 8GB DDR4, Vega10 8GB HBM2, $999, and even that product would(/could) be in a tough spot when a conventially priced PS5 releases.


Why ? Scorpio comes out before ps5 and by how many years we don't know. There could be tens of millions of scorpios sold with just slower zens and gpus attached. It will be much easier to design for scorpio and port down to xboxone/ps4 and up to ps5 . Developers also wouldn't want to give up any installed base to design games exclusively for a new system that sells 0 games. Look at how poor the switches line up is at launch.

Also 7nm wont hit until 2020 for amd if not later. Which gives scorpio 3 years as the most powerful console on the block. A 2020 or 2021 would put the ps5 at the same relative time as a xbox three in my time line. If you have scorpio with zen at 10-20m (or more) ps5 launching and xbox three launching a year later and all 3 have zen plus a gpu why would you give up support for a console that can sell millions of units of your games for one that can sell launch numbers ?

No I think ps5 may find itself in a tough situation on its own right. It wont be as cheap as scorpio or have the install base numbers and it will be eclipsed in power by a new xbox launching a year after it.

Even if you are talking about towards the end of scorpios live in 2026 it will still get games , the games just wont feature the same resolution as they once had. Halo 9 may be 1080p or 1440p instead of 4k and so on. Third party titles will feature the same limitations. But the majority of engines are scalable now
 
I think MS won't force developers to leave behind any Xbox console from now on. Major AAA game developers may stop developing new games for older consoles but Indie game developers will likely keep making new games for them (as long as possible). And even the release of Xbox 4* (I mean Scorpio) and, few years later, Xbox 5 (Scorpio 2) probably won't be the end of Xbox One. Microsoft even has the opportunity to take the advantage of 7nm process and make a new hardware (a set-top box or maybe a hardware with Switch form factor, handheld console, etc.) with current Xbox One APU at higher clock.

*It's a really good time for MS to end it's inappropriate naming system for Xbox consoles. Also it will be a really good way to generate hype for potential buyers and present Scorpio as a next-gen console/hardware.
 
Also 7nm wont hit until 2020 for amd if not later. Which gives scorpio 3 years as the most powerful console on the block. A 2020 or 2021 would put the ps5 at the same relative time as a xbox three in my time line.

For AMD? You mean Global Foundries? AMD uses TSMC to fab Sonys chips. Apple is funding 7nm at TSMC so it will be available as early as 2018, holiday 2019 would be just fine for a 7nm PS5
 
For AMD? You mean Global Foundries? AMD uses TSMC to fab Sonys chips. Apple is funding 7nm at TSMC so it will be available as early as 2018, holiday 2019 would be just fine for a 7nm PS5

For a 150+W, 400mm+^2 chip?

I can't help feeling that optimism and hubris may be encouraging people at this point.

And 7nm still won't allow a density advance that could make a PS5 have the jump over Scorpio that Scorpio has over PS4/XBone.
 
For a 150+W, 400mm+^2 chip?

I can't help feeling that optimism and hubris may be encouraging people at this point.

And 7nm still won't allow a density advance that could make a PS5 have the jump over Scorpio that Scorpio has over PS4/XBone.

7nm offers 3x the density of 16nm and a 65% reduction in power draw

Im not sure why you think Scorpio is even relevant in Sonys roadmap? PS5 needs to offer a big leap over PS4, thats all. And that will be available when 7nm and low cost HBM is available for mass production.. wich is late 2019
 
Status
Not open for further replies.
Back
Top