Xbox One (Durango) Technical hardware investigation

Status
Not open for further replies.
Using 12 separate CUs instead of 3 array of CUs (4 CUs in each array) is the only difference that I noticed, which isn't seems so important.
 
Are you sure? If so then yeah, MS wouldn't really have a legal case to make to take them down.

No, even if that was true (redrawing diagrams), they are still copy and pasting screens from the XDK environment:

xdk1.jpg

xdk5.jpg


Nevermind that most of their articles' written content also looks like direct copy and paste from the ATG docs - the guy behind vgleaks is not a native English speaker and it shows in his other posts where he reposts rumours etc.
 
Last edited by a moderator:
No, even if that was true (redrawing diagrams), they are still copy and pasting screens from the XDK environment:

xdk1.jpg

xdk5.jpg


Nevermind that most of their articles' written content also looks like direct copy and paste from the ATG dos - the guy behind vgleaks is not a native English speaker and it shows in his other posts where he reposts rumours etc.

I don't think he has direct or uninhibited access to XDK as how is the hardware of the alpha and beta kits not important. At least we would get some concrete evidence on what MS thought would be the performance levels of Durango at the time of the alpha kit release. Plus if there was a change in performance in alpha vs. beta wouldn't it be highlighted by MS?
 
Im new to this forum and have been reading post he past month. Is it true that Microsoft upgraded from DDR3 to DDR4. I read this a few pages back.
I had a discussion with friends they said it was no way Microsoft can upgrade becaude of the lack of time to do test the chip.
 
bkilian suggested they did a while back. The diagrams VGLeaks has been using in their articles come directly from MS documentation iirc. bkilian even mentioned how it was interesting that MS wasn't bothering to take the info down.

The diagrams VGleak is using are self made, ergo not actual copyrighted material.
MS can't take them down, they have no legal claim on a self made diagram.

Anyway just because Sony or MS didn't take down the PS4 or Durango specs from VGleaks or DF it doesn't mean that they are false.
Were VGleaks and DF wrong about PS4?
No they weren't and Sony didn't take the specs down.

Really Durango leaked specs (for many) are false because are not confirmed by MS.
 
Last edited by a moderator:
The diagrams VGleak is using are self made, ergo not actual copyrighted material.
MS can't take them down, they have no legal claim on a self made diagram.
No, those diagrams are not made by VGLeaks. Neither is the text written wholly, or even in part, by VGLeaks.
 
What does MS gain from denying the specs, and taking vgleaks to court?

By doing so, they reject their own specs, and in addition, create a negative conotation...not only towards Durango, but the MS brand. We will sue you!...for telling...the truth? How does that resonate with consumers?

Why reject their own product, before it's even announced? Even if MS has legal grounds to have those articles taken down, what would their motivation really be?
 
Most consumers would not care or have a clue if MS sued some obscure blog/website. Hundreds of forum warriors on gaf would surely work themselves into a frenzy.
 
Im new to this forum and have been reading post he past month. Is it true that Microsoft upgraded from DDR3 to DDR4. I read this a few pages back.
I had a discussion with friends they said it was no way Microsoft can upgrade becaude of the lack of time to do test the chip.

Microsoft hasn't shown any willingness to use bleeding edge hardware, so why would they start with RAM?
 
I don't think it's about willingness, it's about what's available for the launch window.
The most reliable rumor has been DDR3 at 2133. Changing it to DDR4-2133 would not be an "upgrade", it's the same speed.
 
Im new to this forum and have been reading post he past month. Is it true that Microsoft upgraded from DDR3 to DDR4. I read this a few pages back.
I had a discussion with friends they said it was no way Microsoft can upgrade becaude of the lack of time to do test the chip.

I suppose it is possible...ddr4 with Rambus R+ for more bandwidth and lower power.
 
No, even if that was true (redrawing diagrams), they are still copy and pasting screens from the XDK environment:


Nevermind that most of their articles' written content also looks like direct copy and paste from the ATG docs - the guy behind vgleaks is not a native English speaker and it shows in his other posts where he reposts rumours etc.

Good catch. Forgot about those. I also have suspected for a while now that they are copy/pasting the diagrams verbatim. There's always been a certain style to them similar to the Yukon leak. Glad bkilian confirmed it. It's interesting why MS isn't going after them then.

I could imagine a scenario where they upped the clocks and decided it would be more beneficial to leverage the leaked info in an effort to surprise ppl with something like a 1.53Tflops GPU and a 2GHz CPU (85% as many flops as PS4's GPU, 25% better clock than PS4's CPU...could be a useful trade off to highlight in marketing).
 
Says who, and even if it were offered, why exactly would they have more to offer?
Point made more of an assumption well spread on the web.
Higher R&D for the on-die SRAM pool is a one-time cost, and pretty much everybody knows how to handle SRAM on-die. SRAM can be engineered for high redundancy, so yield impact is not as bad as raw die area would suggest. There's just one chip on the package, and the SRAM will shrink with any node transition.
True too but I jump to the next point.
A separate eDRAM die like Intel's is sold for maybe an extra $50 for every unit, and it will not integrate with a die with a standard logic process, meaning it's an extra component even if the main APU shrinks.
Well Intel sells it for 50$ because it allows their GPU to breath and may save the laptop manufacturers the cost(s) associated to a discrete GPU. THat is not to say that a 70/80mm^2 of eDRAM would sell for that much. I jump to the next point again.

Are you certain they'd want to use IBM's process?
Well you got me to wonder :LOL: <= sound like a joke but all your previous points were very valid that one is the nail in the coffin.

I didn't not think completely straight on that one, I acknowledge it.
MSFT did not use IBM process for Xenon, Nintendo did with the WiiU though we speak a really tiny chip and not on their most advanced node. As I think more about it, costs might indeed be the reason why Nintendo did not come with an APU but a discrete CPU (which include eDRAM) and a discrete GPU which integrate eDRAM too on a different process (quite possibly cheaper).

I give up on IBM as a solution, you won ;)
I feel like with its return to DRAM business Intel has somehow grown its process advantage significantly.
 
Status
Not open for further replies.
Back
Top