Digital Foundry Article Technical Discussion [2019]

Status
Not open for further replies.
I want to know if the game goes beyond Horizon's GPGPU use for world composition. The huge amounts of ground level detail would be a good fit since the landscape is so devoid of megaflora.

Also, seeing the Death Stranding baby is kind of eerie when I have a two week old baby at home. I hope he's customizable to have the same dark blue eyes as my son ;)

You should play the game with a big jug of orange juice strapped to your waist with your months-old son submerged in it for the full experience.
 
You should play the game with a big jug of orange juice strapped to your waist with your months-old son submerged in it for the full experience.

Does it matter what brand? Would TANG be a suitable replacement?
 
You should play the game with a big jug of orange juice strapped to your waist with your months-old son submerged in it for the full experience.

That would result in murder, unless you can get me breathable liquid a la The Abyss.
 
Large detailed environments were done to death last gen, so reducing graphical assets and detail draw distance to acceptable levels would not be too difficult. It's the CPU side that's more limiting IMO. Alot of the environmental animation and NPCs would need to be culled back. Culled back geometry on the graphical side of things will also help to reduce draw calls.
 
Last edited:
Lockhart is going to be a tough sell among XB1-X and Anaconda.

It will likely depend on availability of games. Lockhart will have a place if the games are not available on the XB1X in the first place. More than 4x the general CPU power (and 8x the SIMD) is a big difference to deal with that might make XB1X porting dubious regardless of GPU spec. But no one will be ignoring the millions of XB1Xs out there so that's an improbable scenario in the first couple years.

Lockhart will have to be marketed directly as a 1080p/psuedo-4k (reconstruction/variable resolution?) experience and MS will have to reiterate it plays the same games and offers the same basic experience as Anaconda that XB1X will not. But what level of GPU capability will it need in order to represent a good fascimile of Anaconda, and what level should Anaconda be at to set a high enough precedent? Telling fans that Lockhart is only 6 TFLOPS just like the XB1X seems a difficult sell, so I'm going to bet MS is going to avoid selling numbers to the public. I can see MS spending alot of time showing games on Lockhart to gain consumer confidence, so I'm hoping that Anaconda is in the 16 TFLOPS range, as to require Lockhart to be beyond the 6 TFLOPS mark to prevent too much downgrades.
 
It will likely depend on availability of games. Lockhart will have a place if the games are not available on the XB1X in the first place. More than 4x the general CPU power (and 8x the SIMD) is a big difference to deal with that might make XB1X porting dubious regardless of GPU spec. But no one will be ignoring the millions of XB1Xs out there so that's an improbable scenario in the first couple years.

Lockhart will have to be marketed directly as a 1080p/psuedo-4k (reconstruction/variable resolution?) experience and MS will have to reiterate it plays the same games and offers the same basic experience as Anaconda that XB1X will not. But what level of GPU capability will it need in order to represent a good fascimile of Anaconda, and what level should Anaconda be at to set a high enough precedent? Telling fans that Lockhart is only 6 TFLOPS just like the XB1X seems a difficult sell, so I'm going to bet MS is going to avoid selling numbers to the public. I can see MS spending alot of time showing games on Lockhart to gain consumer confidence, so I'm hoping that Anaconda is in the 16 TFLOPS range, as to require Lockhart to be beyond the 6 TFLOPS mark to prevent too much downgrades.

No. The DF video is stating how Lockhart will be hard sell if the hardware is around 4TF worth of performance. Making things complicated trying to sell something as "next-generation" when the performance maybe no better than PS4 Pro.

Edit: And 16TF isn't happening. You'll be luck if Xbox-Next or PS5 hits 10TF.
 
No. The DF video is stating how Lockhart will be hard sell if the hardware is around 4TF worth of performance. Making things complicated trying to sell something as "next-generation" when the performance maybe no better than PS4 Pro.
XB1X owners will look at Anaconda as a next-gen machine. XB1/S owners will look at Lockhart and Anaconda as a next-gen machines with different price/performance profiles.
 
XB1X owners will look at Anaconda as a next-gen machine. XB1/S owners will look at Lockhart and Anaconda as a next-gen machines with different price/performance profiles.

I don't disagree with that. I'm just repeating DF's perspective (@7:08) on the matter.
 
No. The DF video is stating how Lockhart will be hard sell if the hardware is around 4TF worth of performance. Making things complicated trying to sell something as "next-generation" when the performance maybe no better than PS4 Pro.

Edit: And 16TF isn't happening. You'll be luck if Xbox-Next or PS5 hits 10TF.

Sorry, I should've written "6 TFLOPS for example". I wasn't directly quoting the DF video. 4/5/6 TFLOPS are all difficult sells IMO. Throw ray tracing into the mix, and you'll be fundamentally cutting down Lockhart at the knees, not because a rastered render won't be good enough to represent Anaconda at a lower res, but because buyers will be missing out on a feature that MS could be pushing up and down as a reason to get a new console.
 
The big Xbox interview - Project Scarlett, cross-gen Halo Infinite and the lack of Fable

Quick opening snip...

You've finally put a name on Project Scarlett and started talking about it. To me, that name sounds like Project Scorpio, a big beefy new console, singular. But last year Phil Spencer talked about new Xbox consoles, plural. What should we expect next year?

Matt Booty: Everything we're talking about - we talked about today. We're not talking about more than what we showed in the video, and what Phil talked about. Project Scarlett is our console plan headed into 2020.

You did say console not consoles plan there.

Matt Booty: Scarlett is our console plan going into 2020. What I'd like to share a little more relevant to my world on the game studios side, is what [Scarlett] will open up in terms of game design. Going all the way back to the early 2000s with streaming, what you do when things are streaming, even some ways Unreal Engine is architected to allow for certain things - going into a room through a little hallway - people have been designing games around load times for so long.

To be able to have the power, the SSD to unlock a new approach to game design... We're probably sitting on a pivot point in game design, when you add up the new console's speed and performance, what cloud streaming will offer up when all the instances of a multiplayer game are running in one location, what it means to be running a game in a data centre... all that in combination with some of the hardware stuff could be as big a transition as when we went from 2D to 3D.

Think when the first 3D console first showed up. A lot of games were 2.5D. It took a while for people to figure out, y'know, how do you make a fighting game in 3D? How do platformers work? I feel we're sitting on a pivot point equal to that because of what the hardware will unlock.
 
Back to Scarlett... you mentioned it has four times the power of Xbox One X, which certainly sounds good. But what does that mean?

Matt Booty:
It's a few things - it's the combination of speed, not just of the SSD but of the processor, the performance of the GPU and RAM, but we're also in a world where speed is starting not to matter. You can make RAM faster either by speeding up the way you access it or by adding more access points. Just think, what are all the things right now which take you out of a game? You're playing then suddenly *bloop* a load screen pops up and drops you out. Our goal is to get rid of those things, that's what we're after.
 
I was playing portal 2 yesterday and I was shocked at the loading screens like I looked online to see if the game was configured wrong or something because I totally forgot that the game even had them.

Totally disrupts an otherwise amazing game.
 
I was playing portal 2 yesterday and I was shocked at the loading screens like I looked online to see if the game was configured wrong or something because I totally forgot that the game even had them.

Totally disrupts an otherwise amazing game.

Indeed. It was such a bother, even I complained about it previously...


1) Too many loading screens.
2) Some textures look like the same ass from 2008.
1) Too many loading screens.
3) Way too many jaggies.
1) Too many loading screens.
4) Inability to pick more precise replay location other than chapter starting points.
1) Too many loading screens.

Portal 2 has entirely way too many loading screens when compared to console games. That is not an issue caused or limited by consoles. It is a limitation on how Valve coded it up.
 
Indeed. It was such a bother, even I complained about it previously...
Lol I guess the first time around they were still pretty commonplace so I thought nothing of them. This time it really pulled me out of the experience.
 
The big Xbox interview - Project Scarlett, cross-gen Halo Infinite and the lack of Fable

Quick opening snip...

You've finally put a name on Project Scarlett and started talking about it. To me, that name sounds like Project Scorpio, a big beefy new console, singular. But last year Phil Spencer talked about new Xbox consoles, plural. What should we expect next year?

Matt Booty: Everything we're talking about - we talked about today. We're not talking about more than what we showed in the video, and what Phil talked about. Project Scarlett is our console plan headed into 2020.

You did say console not consoles plan there.

Matt Booty: Scarlett is our console plan going into 2020. What I'd like to share a little more relevant to my world on the game studios side, is what [Scarlett] will open up in terms of game design. Going all the way back to the early 2000s with streaming, what you do when things are streaming, even some ways Unreal Engine is architected to allow for certain things - going into a room through a little hallway - people have been designing games around load times for so long.

To be able to have the power, the SSD to unlock a new approach to game design... We're probably sitting on a pivot point in game design, when you add up the new console's speed and performance, what cloud streaming will offer up when all the instances of a multiplayer game are running in one location, what it means to be running a game in a data centre... all that in combination with some of the hardware stuff could be as big a transition as when we went from 2D to 3D.

Think when the first 3D console first showed up. A lot of games were 2.5D. It took a while for people to figure out, y'know, how do you make a fighting game in 3D? How do platformers work? I feel we're sitting on a pivot point equal to that because of what the hardware will unlock.

What I took from that.

Scarlett is their console plan. He's being very precise about wording because he doesn't want to say too much and at the same time he doesn't want to be caught in a lie. That's their plan on what they are going to do with next gen.

That doesn't mean that there aren't multiple Project Scarlett consoles. There might be 3 or 4 of them for all we know. Anaconda, Lockhart, Streaming, etc.

So Phil talking about Project Scarlett consoles doesn't necessarily contradict what Matt Booty said in the interview.

Their Scarlett Plan likely encompasses more than just the console hardware.

Regards,
SB
 
Last edited:
Status
Not open for further replies.
Back
Top