Xbox One (Durango) Technical hardware investigation

Status
Not open for further replies.
XboxOneDev on Reddit made this post in regard to background downloading in low power mode...



http://www.reddit.com/r/xboxone/com...eally_cool_feature_will_the/cbwvsgq?context=3

Tommy McClain
It would be nice to have hybrid state also.
Where every half hour it moves from lowest state, to higher and checks online for any updates etc, then goes back to lowest state.
It would only need to be in higher state for few seconds to check, or obviously longer if there was something to download.

Would also be nice if you could set it to go from lowest state to the one where you can say "xbox on", at a specific time, so you can set it knowing roughly normal time you get in from work, etc.
That way it can be in lowest state all day.
 
It would be nice to have hybrid state also.
Where every half hour it moves from lowest state, to higher and checks online for any updates etc, then goes back to lowest state.
It would only need to be in higher state for few seconds to check, or obviously longer if there was something to download.

Would also be nice if you could set it to go from lowest state to the one where you can say "xbox on", at a specific time, so you can set it knowing roughly normal time you get in from work, etc.
That way it can be in lowest state all day.

Good suggestions and I'm sure MS is working on a software up for those options.:smile:
 
Sorry for not making it clear before that he works at DICE, but sebbbi is very technologically minded and he is probably working in the game.

Anyway, all I know is that despite his saying he is one of the main programmers of Trials games, he is also involved in other games afaik.

repi works for EA DICE. sebbbi works for RedLynx.
 
repi works for EA DICE. sebbbi works for RedLynx.
My bad, it seemed evidently clear to me that he worked at DICE, which was one of the main reasons why I wanted to try BF3 in the first place. I am still wondering why I was generally clueless as to where he works. I will try to fix my previous post.

I knew that repi is one of the developers of Battlefield 4 though.
 
....except the ESRAM is not a cache it's a scratchpad. I haven't seen any indication of cache logic in the diagrams.
Oh, okay. That doesn't sound like something I'd say. What's the difference between a software cache and a scratchpad? Does it have any implication in the framebuffer size?
 
I knew that repi is one of the developers of Battlefield 4 though.
He's not one of the devs on BF4, he's technical director for Frostbite engine. ;)

Oh, okay. That doesn't sound like something I'd say. What's the difference between a software cache and a scratchpad? Does it have any implication in the framebuffer size?
Cache is part of the memory hierarchy and keeps an easy to access the *copy* of the data that's stored somewhere in another (slower) memory. Scratchpad is directly addressable* and can/will contain the only copy of some data. E.g. for deferred renderer you build your g-buffer and then use it as a source for next render to build the final composite. There's no point in pushing g-buffer to RAM if you can keep it temporarily in a large enough and pretty fast piece of memory. This is what you'd use your scratchpad for.

* embedded memory in X1 is behind the MMU according to diagrams so it's probably not directly addressable but mappable into the VA space
 
I am more curious about the design of the 2 gfx cmd proc + 2 compute cmd proc for the Xbox One...it seems that PS4 uses 1 general cmd proc with 8 ACEs (normally 2), anyone can shed some light on this?
 
have not seen it either but
its certainly referenced in the extremetech article.

The other major mystery of the ESRAM cache is the single arrow running from the CPU cache linkage down to the GPU-ESRAM bus. It’s the only skinny black arrow in the entire presentation and its use is still unclear. It implies that there’s a way for the CPU to snoop the contents of ESRAM, but there’s no mention of why that capability isn’t already provided for on the Onion/Garlic buses and it’s not clear why they’d represent this option with a tiny black arrow rather than a fat bandwidth pipe.
 
its certainly referenced in the extremetech article.
It's on none of the slides, though. :S IDK, but perhaps SRAM can be mapped to the CPU VA space? This would probably use up GPU DRAM BW though so it wouldn't be very useful and severely BW limited compared to the SRAM<->GPU throughput.

We are talking about a +300% exaggeration of 'Jaguar -to- memory' bandwidth
ORLY?
http://www.intel.com/content/dam/doc/product-brief/x79-express-chipset-brief.pdf
12,8 x 4 = 51,2 < 68
Drawf is not the right word but your 300% comes from nowhere.
 
It's on none of the slides, though. :S IDK, but perhaps SRAM can be mapped to the CPU VA space? This would probably use up GPU DRAM BW though so it wouldn't be very useful and severely BW limited compared to the SRAM<->GPU throughput.


ORLY?
http://www.intel.com/content/dam/doc/product-brief/x79-express-chipset-brief.pdf
12,8 x 4 = 51,2 < 68
Drawf is not the right word but your 300% comes from nowhere.

the jaguar only has 22.4 GB/sec communication, NOT 68GB/sec. Do the math.
Disclaimer: 300% is an estimate, it possible that it's 298.9% or 301.2%
 
the jaguar only has 22.4 GB/sec communication, NOT 68GB/sec. Do the math.
Read the entire sentence, please. "Jaguar in the Xbox One" is not Jaguar in your PC (and on top of that calling GPU in Xbox One "Jaguar" is probably a bad idea to begin with). Non-CPU-cache-coherent GPU bandwidth on X1 is 68GBps.

Oh, cool, thanks! :) Its placement is super weird. :S
 
Status
Not open for further replies.
Back
Top