Which is a massive downside.... unfortunately.
As for the 880GB/s... i wouldnt discard it. However it might just be the dev kit has 2 Vega GPU running at 440GB/s.
-
Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.
-
We have made minor adjustments to how the search bar works on ResetEra. You can read about the changes here.
PS5 and next Xbox launch speculation - Post E3 2018
- Thread starter Phoenix Splash
- Start date
Which is a massive downside.... unfortunately.
As for the 880GB/s... i wouldnt discard it. However it might just be the dev kit has 2 Vega GPU running at 440GB/s.
So looking at costs more in-depth, it would seem that HBM is not happening from either player in the next gen of consoles, unless they're willing to eat significant losses.
I imagine we're looking at GDDR6 on a 384-bit bus from both? What would the bandwidth figure be there?
Which is a massive downside.... unfortunately.
As for the 880GB/s... i wouldnt discard it. However it might just be the dev kit has 2 Vega GPU running at 440GB/s.
Depends on the speed...
A 384-bit bus at 12gbps -> 576GB/s
at 14gbps -> 672
at 16gbps -> 768
In a console, i very much doubt you will get 16gbps ram. Which means at most you will get 14. So 672GB/s is the most likely using a 384-bit bus. Or 448GB/s using a 256-bit bus. However i'm very skeptical Sony (or Microsoft) using a 384-bit bus in their next-gen console (at least for their base PS5 and Scarlett).
Hybrid....cube memory? I'm just spitballing, i'm assuming it would be more expensive than even HBM. I would like to point out if we had Krazy Ken back we would probably end up with HCM lol
What would the advantage of having the cooling on the underside of the motherboard even be? It seems needlessly complicated
All in good fun, mate. It was more of a playful dig.
HBM IS stacked DDR memory.
That patent is potentially exciting.
So it might be for some type of 2.5D or 3D stacking, be it processor stacking and/or memory stacking (ala HBM) within a chip package, correct?
So it might be for some type of 2.5D or 3D stacking, be it processor stacking and/or memory stacking (ala HBM) within a chip package, correct?
Fair enough, was not aware, thanks for the info. Would their be any advantage to a cooler on the underside of mobo if not for memory cooling on the opposite side?
So this is interesting, but does this picture look familiar?
Now obvsiouly Sony is not working with intel, just thought it interesting. This pic is from an article released hours ago.
https://www.top500.org/news/intel-goes-vertical-will-stack-logic-chips-into-3d-packages/
I believe that's what Scorpio uses already. I expect XBOX to go beyond that. They hired a next gen memory person not too long ago.
I wouldn't think so.
The transistors are where heat is generated. Placing a cooler on the mobo underneath would be placing an additional barrier of thermal resistance between the heat source and the heat sink. You want your heat sink mounted directly on the chip packages, to pull away as much heat as possible and keep those cores running efficiently.
Xbox uses GDDR5 on a 384-bit bus.
A question though: How would MS fair with backwards compatibility in regards to using a 256-bit bus with GDDR6? Or am I complicating things and the only thing that actually matters is raw bandwidth numbers?
I went back and checked the pdf, it says it allows for increased degree of freedom when arranging components. My understanding is laymen of course, but i also assumed that having the cooler on the underside makes it less in cooling. Is any of this making sense? Maybe we're missing a piece and really they have found a way to make it just as efficient to cool from the underside. Or they are spending extra on cooling to compensate and they feel it will give them enough of an advantage for other components in how they're arranged
You're right, that was GGDR5.
I dont think it matters much as long as the throughput is there.
HBM is only higher bandwidth if they use enough stacks or somehow get HBM3 in there. Otherwise, GDDR6 can beat up to 3 stacks with a 384 bit interface.
Yes. Or they could even stack chiplets if they had TSVs. That's pretty advanced, though.
So this is interesting, but does this picture look familiar?
Now obvsiouly Sony is not working with intel, just thought it interesting. This pic is from an article released hours ago.
https://www.top500.org/news/intel-goes-vertical-will-stack-logic-chips-into-3d-packages/
That's 2.5D stacking. To be fair, it's kind of a gray area given that's an active interposer, but the logic die themselves don't have TSV, which is the traditional threshold for '3D' stacking. The patent image shows potential 3D stacking.
There would not solely be a motherboard interface. The image shows some kind of metal interface on the top of the PWB, which interfaces to through-hole metal posts, which then interface to a true heatsink on the opposing side of the PWB. You'd be more thermally limited than a conventional heatsink (not all of it could be air-cooled, you create a non-ideal metal interface, etc.)
However, you could potentially mount your die upside down on the underside of the package, still allowing for direct die contact with your thermal interface. In fact, I think this is exactly what is depicted with 5a. Then, you could stack your memory or other chiplet device on the other side of the package. This way, you achieve a 3D-like solution without the need for any TSVs potentially (depending on the density of interconnects achieved in the package). Your chips certainly wouldn't need TSVs at least, unless they are stacked on themselves like HBM.
Additionally, some packaging technologies allow for embedded active components. Again, this is another exotic, probably more far-out technology, but if the dissipation is low enough, one could certainly do it.
Only bandwidth matters. The memory controller is going to abstract the particularizes when it comes to how the CPU and GPU get their bandwidth. GDDR6 is actually more granular than GDDR5, so there's no worries there.
Last edited:
Do you think there are spies in both camps that divulge info to both MS and Sony. In other words, One knows what the other is doing right?
Pffft. Codename is clearly "penis". Any insider who says thats the name no on will believe and any source that tells someone the codename is "penis" will just be sneered at and told to screw off. It's absolutely GENIUS
Would also make for really easy and comprehensible internal communication. They just have to make sure they refer to the actual PS4 as something else like, dunno, PS3.
Guys....it's "Penis", let's be real here
Real talk here, Microsoft is on their 7nm devkit after a 16nm version. How fricken early did they lock down those specs?
What makes you think early devkits are not just PCs?
First Durango devkits were Intel Xeons with 7800 series gpus.
I know from first hand experience that Zebra Durango devkits arrived internally in Spring of 2013 at Microsoft. We're more than a year out from getting final Scarlet silicon.
To be honest, I would hope he is this time. Unless they just decided on a Zen 2 and hyper souped up vega with a fancy cooling solution. Which would be interesting against a supposed Navi
Edit: relying on the cooling solution to upp everything would kind of fit with the 1x. It wasn't anymore exoticly custom the can og x1 or ps4. They just had a good cooling solution and amped everything up
But supposedly these aren't the first devkits, these are the second. The first round were on 16nm, these are 7nm. You already have the die shrink, why bother with a third iteration?
Where is the source for 7nm? If we have 7nm devkits then Scarlet is launching in 2019.
Considering the X1 has esram and ddr3 and th 1X has gddr5, I think Ms will be just fine in the BC department
MS going to design Scarlet on 16nm and then waste money to port to 7nm obviously! Also pay premium for 7nm chips two years before launch just so that they can boast that they have devkit that looks like the final thing way ahead of Sony.
I guarantee you that unless the console is releasing in 2019, any dev kit right now is just a PowerPoint recommending the general ball park power, and devs can use Intel or AMD or Nvidia.
Chuckwalla was previous devkit, current devkit is anaconda. It's there second devkit for next gen
Jez only leaked the DevKit names and not that there are real 7nm console dev kits are already out there. PC dev kits usually are spec'ed on the target and not on the finalized specs.
Okay okay, I'm man enough to admit when I've made a mistake about the devkits lol. That being said, this is either controlled leaks by Microsoft or Sony has everyone with any info hooked up to a shock collar and GPS
Let me preface this by saying their is nothing wrong with it, journalism is a hard job. But Jez is clearly playing the role of the member in the media who a company uses to release info without actually releasing info. In sports for example, it's a suuuuper common thing that an organization will use a reporter in that role. Whether it be to weed something out, kick up a rumour for their own benefit or soften the blow of something. Again, there's nothing wrong with it, no one loses. It's a win win, we get info, Jez gets his info to release and MS gets the hype machine rolling
IMO I don't have issues with controlled leaks as long as it isn't deceptive; it becomes just another way to market and/or release info without needing to do a full press release.
I would say there is just as much chance in it being an organic leak.
What you have said does not really suggest its a controlled leak.
I mean you could say what you just said about any leak.
Did I miss a bunch of MS leaks or something? I thought all we had was a codename for the dev kit and that it's probably using Zen 2?
That's the exact role that Jez is being "used" for. And I put used in quotations because none of it as you said is lies, but I'm sure he knows the info he gets is to Microsofts benefit.
Yes you are right, I'm just going off my general experience with sports for my comparison point though. Nine times out of ten when a team has a reporter or two that get literally all the inside info over a significant period of time, a quid pro quo is going on.
We got a retroactive leak funnily enough, the name of the 1x devkit which turns out was in an article of the daily Star a while back. And then at like the same time someone on the board found an an overly detailed recent Sony patent for a funky cooling system which COULD indicate a 3d stacked ram situation
I am content, with it being a gut feeling.
That makes sense to me.