Predict: The Next Generation Console Tech

Status
Not open for further replies.
Okay, things in the article:

-(Mag speculation) Will just be called Xbox, not XB3, XB720 etc. Like Apple's iPad.

-Claim everything else in the article is sourced, not speculation

-Durango titles mentioned being worked on (not sure if this is supposed to be any new info or just gleaned from the internet) Kinect Sports 3, Bungie's Destiny, Fable MMO, Halo 5, Battlefield 4, Forza 5.

-Claim "unless something changes" Durango will be revealed at Microsoft's X showcase in February or March (though bkilian seems to disagree)

-Claim rumors say the controller is prototype stage, could feature a touchscreen or programmable buttons.

-Interesting claim that their could be a "core unit" without an optical drive. Kind of a new definition of the core unit I guess.

That's it that hasn't already been mentioned. Mostly generic fluff in the article. Things like an on record Nvidia engineer (note, this has no implication Nvidia is building the GPU) saying "WELL I THINK ITS SAFE TO SAY THE GPU IN ANY NEXT GEN XBOX WOULD LIKELY FEATURE DX11 FUNCTIONALITY WHICH WOULD ALLOW BULLET POINTS X Y AND Z HYPOTHETICALLY". Yawn.
 
-Durango titles mentioned being worked on (not sure if this is supposed to be any new info or just gleaned from the internet) Kinect Sports 3, Bungie's Destiny, Fable MMO, Halo 5, Battlefield 4, Forza 5.

Lionhead MMO? I guess there is not "fable MMO" in development, it is a new IP from Lionhead (the fable makers).
 
"My ex-girlfriends' former roommate's nephew's brother in law's veterinarian's IT department head said that Durango will feature 800 quadrillion bits and a total linear computational speed of 60 trillion operations per second. So it must be true."
Damn, that dude is in the _know_. My dog's chiropractor's fiance's other brother Darryl agrees with him.
-Claim "unless something changes" Durango will be revealed at Microsoft's X showcase in February or March (though bkilian seems to disagree)
The 360 reveal was at E3 (well, just before) for launch the same year. Why would MS want to tank their 360 sales for 3 quarters if they didn't have to? It makes no sense to reveal new hardware with a long lead time, marketing wise it's a nightmare, sales wise, it's unwise, and historically, they haven't done it.
 
-(Mag speculation) Will just be called Xbox, not XB3, XB720 etc. Like Apple's iPad.

Unlikely.

-Claim everything else in the article is sourced, not speculation.

Unlikely.

-Durango titles mentioned being worked on (not sure if this is supposed to be any new info or just gleaned from the internet) Kinect Sports 3, Bungie's Destiny, Fable MMO, Halo 5, Battlefield 4, Forza 5.

Anyone could have predicted those titles. Suprised Alan Wake 2 isn't listed. :rolleyes:

-Claim "unless something changes" Durango will be revealed at Microsoft's X showcase in February or March (though bkilian seems to disagree).

"unless something changes" :LOL:

-Claim rumors say the controller is prototype stage, could feature a touchscreen or programmable buttons.

Which means chances are that it wouldnt'.

-Interesting claim that their could be a "core unit" without an optical drive. Kind of a new definition of the core unit I guess.

Again, "could"

That's it that hasn't already been mentioned. Mostly generic fluff in the article. Things like an on record Nvidia engineer (note, this has no implication Nvidia is building the GPU) saying "WELL I THINK ITS SAFE TO SAY THE GPU IN ANY NEXT GEN XBOX WOULD LIKELY FEATURE DX11 FUNCTIONALITY WHICH WOULD ALLOW BULLET POINTS X Y AND Z HYPOTHETICALLY". Yawn.

Conclusion? Turrible attempt to sale copies.
 
Which means chances are that it wouldnt'.

I agree. For sure I am positive MS wont attempt a full on mimic of the Wii U controller with all it's drawbacks. And I'm not sure what good "programmable buttons" would be. But a small touchscreen in the controller might be worth exploring, not that I'm predicting it will happen. I had though that the 360 controller was probably about as good as controllers could get, but a (small) touchscreen in the center of the controller sounds intriguing.

Again, "could"

Again, interesting concept that might be worth exploring. At first glance optical drives are cheap, while hard drives are relatively expensive. So leaving out the optical drive would seem to have little benefit. But, when you factor in the form factor, cooling, reliability (all proxies for cost as well), etc benefits that could be gained, it might be more appealing. Again not that I expect this to actually happen. It would also seem to depend on a wholesale change by MS to day and date DD, else it's unfeasible off the top.

Conclusion? Turrible attempt to sale copies.

Mostly yes. "8 page spread" is 90% graphics and empty space and 10% text. Of that text even less isn't generic fluff. The only interesting specific info is the 16 cores/8GB.
 
Lionhead MMO? I guess there is not "fable MMO" in development, it is a new IP from Lionhead (the fable makers).

In either case, no wonder Molineux wanted out. Every MMO released in the last few years that wasn't Korean has been an expensive misadventure.
 
In either case, no wonder Molineux wanted out. Every MMO released in the last few years that wasn't Korean has been an expensive misadventure.

But it is not a MMO, it is "like" a MMO:

"This is to assist in the development of a major brand new IP that is being targeted for the future generation of platforms and will be an RPG based game with a radical new take on how co-operative and multiplayer gameplay feeds into the experience, while blending online and single-player into one complete experience."
 
If we factor in, that 4 Jaguar cores form a "module" with a common L2 cache, it could mean 4 "modules" or 16 Jaguar cores. That would be more computing power than I had expected (but one would need a lot of threads to use it efficiently). And it would somehow fit to the first devkits allegedly containing two quadcore Xeons (16 threads with HT). I don't know.

I'm not sure putting down 16 Jaguar cores makes the most sense, but if they're only 3.1mm/core as AMD said. 16 would fit in 50mm. You're probably looking at another ~25 for the L2 caches and would probably need an L3 cache plus all the glue logic ~50mm. That CPU could be around 125mm which is still smaller than the original 360's initial CPU area. It would probably be under 40W as well.

I do wonder if we would get some sort of custom SMT Module . If you look at the layout of the a Jaguar core it looks like %50 of the area is the caches + FP unit. Perhaps they can tweak the design to share some of those components and make a 4 core module.
 
I don't think this could then be called Jaguar, and that could be horrendously expensive and complex, sort of creating a mini-Bulldozer. AMD would surely only afford to maintain the Jaguar and Bulldozers designs, but not a third one.

I'd go as far as to suggest the four Jaguars are independant. No L3, little glue : you effectively have four islands, even less connected on a logical level than multi-socket Opteron and Xeon. No cache coherency, it's only done at each of the four shared L2 caches.
This is at least the arrangement that can be expected on AMD's ARM Opteron, where there might be, who knows, eight multi-core ARM CPUs on the die.

This would make the chip "supercomputer-like" in a sense (whereas I wouldn't say the same if the CPU would be regular SMP). All the HSA stuff would work between the four "CPU islands", like between a CPU and the GPU, so you have at the very least a unified addressing space. Crazy bandwith and low latency too.
 
is it possible to put hd 8770 into amd a10 apu - either having piledriver cores or steamroller(which is delayed again) ? it would be close to the ps4 spec which leaked on vgeaks !
 
I'd go as far as to suggest the four Jaguars are independant. No L3, little glue : you effectively have four islands, even less connected on a logical level than multi-socket Opteron and Xeon. No cache coherency, it's only done at each of the four shared L2 caches.

I think this is extremely unlikely. From what we know about Jaguar, it's probably perfectly capable of doing glueless multiprocessing with 4 nodes. Doesn't need a L3 to do that.
 
Oh, I may be ridiculously wrong but I thought I was clever when I wrote this.
I wondered if there are some significant costs (albeit maybe small) in power or design complexity. It's Xeon Phi leaving out L2 coherency, and presumably AMD's ARM, that made me think this.

I also thought that in Jaguar's core markets, low end desktops and laptops, high end tablets, there's no need at all for multi-CPU operation (whereas all other AMD architectures except Bobcat targetted multi-CPU servers as well as desktops/laptops).
AMD could have thus left out all consideration about it, and did copy-pasta their design unmodified in a console. But maybe this reasoning is weak.

I can agree there's no L3, no matter what.
 
is it possible to put hd 8770 into amd a10 apu - either having piledriver cores or steamroller(which is delayed again) ? it would be close to the ps4 spec which leaked on vgeaks !
I'm not an expert by any mean but sadly if AMD could pull this I would expect to do it first for them selves, they need it. As it is the GPU they integrate in their APU lag their discrete GPU production.
 
I can agree there's no L3, no matter what.

Thinking about it more, I think you're right. With no L3 cache and no memory coherency between clusters, the design is much simpler. With hardware coherency, you'd have a lot of hardware for the address checking among all the caches. Shifting the burden to the programmer is probably a better solution for a power/area efficient design. Even with 16-cores, it would be a really small design, probably sub ~100mm.
 
I'm not an expert by any mean but sadly if AMD could pull this I would expect to do it first for them selves, they need it. As it is the GPU they integrate in their APU lag their discrete GPU production.

May be it is exclusive to sony or amd will launch such type of apu after the ps4 been launched!
 
I'm not an expert by any mean but sadly if AMD could pull this I would expect to do it first for them selves, they need it. As it is the GPU they integrate in their APU lag their discrete GPU production.

Yes and no. I don't think the problem is so much putting a fast GPU with a CPU, it's memory bandwidth. Even fast 2.133 DDR3 isn't cutting it. It's only about 17 GB/s 64-bit channel and I think the APU's only support dual channel as well, but even four channels wouldn't be enough. And it's not like you can buy off the shelf GDDR5 modules, so AMD is kind of stuck. I think embedded RAM will be used down the road.
 
IBM quad core would more likely be POWER8 based in my opinion.
I wish that would be true, power 8 or a custom power 7+ on IBM 32nm process.
Actually the overall design for the 360 proved to be quiet successful, the only draw back was the Edram which remained unchanged at 10MB when MSFT decided to move to 720p /HD resolution as the target resolution for the system.

It could be achievable to deliver the jump in performances people wants while actually possibly lowering a tad the silicon budget vs the 360.
I'm not sure about how big would be a quad core power 7+ with a sane amount of L3 (not 10MB per core) vs Xenon. Looking at a power 7+ die shot as this one they look indeed pretty tiny.
Then you have the massive L3 interconnect, the SMP links, memory controllers, accelerators, etc.

I would think that a "ROPless" HD 7850 (16 CUs native) with let say three memory channel /192 bit bus to DDR3 would end in the same ball park as Xenon. Pitcairn is 202 mm^2 and CUs seems to take a lot of that space (Anandtech among others has die shots available here ).
In a "native" HD 7850" (as I describe it), that is 20% less space allocated to the CUs, no ROPs, 25% less space taken by the memory controllers, I would not be too surprised if the chip ends up at least 20 % tinier than pitcairn, that would be a max estimate of 200 mm^2 minus 40mm^2 so 160 mm^2.

That left the smart edram, using IBM 32nm process, too I would not expect 64MB to be that big.
 
Status
Not open for further replies.
Back
Top