Next Generation Hardware Speculation with a Technical Spin [2018]

Status
Not open for further replies.
the xbox one x is already small tho

you can always go smaller. The xbox one x is a very powerful console and it could be a really smart move for MS to cost reduce it and start transitioning it a low end console for a 2020/21 launch of a new console. The 6tflop machine should be able to run the next gen game at lower resolutions decently. Games designed for a 4k next gen machine should be able to run on the x at 1080p with reduced quality settings. For ms having a $250 ish console wouldn't be a bad idea

I would think that's a branding nightmare, I would assume they cost reduce but no name change or fanfare. I am not sure I see a redesign either, it's already small. Reducing the complexity of the cooling and hovis method as well as a cheaper Apu would be good for them. if they shrink the case it probably errodes these savings when the console is already small.

I mean the name could be anything . I was just using it as a way of describing it. A 7nm APU should give them some breathing room with cooling and costs.

I am also sure that at some point in the near future it will be very hard to cost reduce the xbox one nad ps4 much further . So if it comes towards the end of 2019 it could be the only xbox still for sale at that point
 
Last edited by a moderator:
What are the chances of this PS5 Navi hitting 15 tf now that we know AMD is pouring their life blood into it and maybe trying for something more than just a mid level gpu?
 

Allegedly, Koduri saw up to a massive 2/3 of his engineering team devoted exclusively to Navi against his wishes, which resulted in a final RX Vega product Koduri was displeased with as resources and engineering hours were much lower than anticipated. As I mention in my companion report, the implication is that AMD CEO Dr. Lisa Su wanted to devote more energy into its semi-custom business than its desktop graphics division.

Parts of the Forbes story has a ring of truth to it. Mostly about the internal fighting between Raja Koduri and AMD's CEO Lisa Su. It seems from Kyle Bennett 2016 story, Koduri had serious issues with Su's leadership direction, where AMD was headed, and the redistribution of resources from other projects (i.e., Polaris, possibly others). And he also wanted AMD to spin off the GPU side into "ATI" again. So yeah, he's no longer there...
 
What are the chances of this PS5 Navi hitting 15 tf now that we know AMD is pouring their life blood into it and maybe trying for something more than just a mid level gpu?
FP32? Well SoCs don’t run on lifeblood, luckily, so the performance will mostly depend on lithographic process and power draw envelope.
Lisa Su isn’t performing unholy blood sacrifices (well, maybe she is, but not to get the chips to run faster, but possibly to get them out the door faster!) so nothing much has changed in terms of prognosis. With sufficient power budget for the device 15TF (FP32) is clearly possible.
 
I'm going to go all in clueless and say that the next xbox is going to use an Intel chip. It'll be an APU but using their 2020 dGPU architecture.

Intel's interest is that it's a multi billion Azure deal. The next Xbox is as much an Azure product with
MS's Xbox everywhere stance.

When you play Xbox cloud I don't think it's unreasonable to assume it could be running on any Win10/DX12 capable hardware with sufficient grunt.

Don't really think Intel will happen, but it does feel like the Xbox team could use whichever vendors they like and achieve backwards and forwards compatibility.
 
I would be shocked if Xbox went with Intel. It would kind of makes sense though if the 2/3 of AMD engineers working on navi for Sony is true. Would be a win for Intel though as in getting at least 40 million plus chips out there in the Xbox alone. Would it mean a Intel CPU also?

Would make the next gen discussion much more interesting.
 
Console apu/cpu busines is usually low profit I think, even if they sell millions, but with minimal margins.

So intel with their over priced cpus with big margins maybe doesnt care much about console world.
 
Console apu/cpu busines is usually low profit I think, even if they sell millions, but with minimal margins.

So intel with their over priced cpus with big margins maybe doesnt care much about console world.
Yeah but for a graphics chip, they did announce a dedicated graphics card coming in 2020.
 
What are the chances of this PS5 Navi hitting 15 tf now that we know AMD is pouring their life blood into it and maybe trying for something more than just a mid level gpu?

Personally I wouldn’t be surprised. I think Vega has a lot of potential for power optimization. It will be interesting to seethe performance/watt of Vega 7nm to see how the architecture has scaled to the new process.

Navi should increase that ratio much higher.
 
What better way to make a splash with their dGPU tech than powering the next Xbox?

Still think it's unlikely, but their 2020 GPU strategy makes it more plausible.
Yeah and it kind of makes sense with the Navi Sony news. This is getting a bit fairy tale ish :) would be brilliant for discussions though Intel vs AMD:yes:
 
FP32? Well SoCs don’t run on lifeblood, luckily, so the performance will mostly depend on lithographic process and power draw envelope.
Lisa Su isn’t performing unholy blood sacrifices (well, maybe she is, but not to get the chips to run faster, but possibly to get them out the door faster!) so nothing much has changed in terms of prognosis. With sufficient power budget for the device 15TF (FP32) is clearly possible.
So no secret lifeblood sauce then:LOL:? I totally don't mind a bigger console box opting for 200w if we can hit the numbers, there's always the Slim version for later.
Personally I wouldn’t be surprised. I think Vega has a lot of potential for power optimization. It will be interesting to seethe performance/watt of Vega 7nm to see how the architecture has scaled to the new process.

Navi should increase that ratio much higher.
That's good to hear!
 
There are three things that make me think Microsoft could go in an interesting direction next generation: this talk of Microsoft using Intel, the recent rumour of the XBoxOne successor being a family of devices, and the often proffered idea of the X1X playing next gen games.

What if Microsoft's next generation offering is a platform? I'll call it "XBox Infinite" because of the recent Halo announcement. Their next generation games are branded with this, as is the X1X, which, over time, can become their 1080p30 machine.

They can use whichever vendors they like, akin to Steam machines, and release devices across different prices and performance levels. Ideally, denoting performance:
- Tier 1-
1080p30 minimum. X1X performance or higher.
- Tier 2-
4KCB60 minimum (or some alternative form of reconstruction). Double X1X performance or higher. Still within the scope of an SoC.
- Tier 3-
4K60 minimum. High end gaming PC performance, necessitating an MCM design, a large case, lots of cooling, and a high price.

They could have the most powerful console at the highest price, an SoC based console to compete on price with the PS5, and a budget console with an already established market, mitigating some of the risk of launching a new generation.
 
What better way to make a splash with their dGPU tech than powering the next Xbox?

Still think it's unlikely, but their 2020 GPU strategy makes it more plausible.

I see it as a way to force developers to optimize to their architecture, which is then a boon to PC performance and helps increase adoption. Intel was reportedly trying to sell Larrabee to console makers.

I’d bet they’re willing to make some concessions on margin to get a foothold in the console space. If they can finally get 10nm healthy by 2020, it could be a compelling offering. Still, I’m skeptical their first effort will be performance competitive.
 
Intel in Next Xbox seems unlikely but Intel doing something interesting in the graphics space is only a matter of time. Intel are literally throwing every technology and combination of conventional and unconventional technology at the wall and seeing what sticks. And they desperately need something to stick because their main business of selling volume x86 and x86 chipsets is diminishing.
 
What are the chances of this PS5 Navi hitting 15 tf now that we know AMD is pouring their life blood into it and maybe trying for something more than just a mid level gpu?

It’s not simply being Navi. From semiconductors articles I read on the web (I’ll post links when I get a chance), a midrange 16 nm chip needs about 200 engineer years to produce a working chip. For 7nm that number jumps to 500 years. You literally need 2.5X as many engineers at 7nm than you need at 16nm to meet a similar time frame for release.

So at 16 nm, AMD had to source resources for Vega, its PC derivatives and two mid gen console APUs mostly based on Polaris, an existing IP. For 7nm you have Navi with PC derivatives plus two next gen consoles with much higher potential for sales than mid gen consoles.

From a business perspective it makes a ton of sense to have a much larger design team for 7nm versus 16nm.
 
Last edited:
Intel in Next Xbox seems unlikely but Intel doing something interesting in the graphics space is only a matter of time. Intel are literally throwing every technology and combination of conventional and unconventional technology at the wall and seeing what sticks. And they desperately need something to stick because their main business of selling volume x86 and x86 chipsets is diminishing.

Agreed. Given the markets that exist for gpus now, it only makes sense for Intel to go all in. But for console manufacturers it would be a giant risk to choose a manufacturer that just been concentrating on “good enough” for the past 30 years.
 
Status
Not open for further replies.
Back
Top