Monday, June 13th 2016
Microsoft XBOX Scorpio SoC Powered by "Polaris" and "Zen"
It looks like Microsoft will overpower Sony in the next round of the console wars, with a more powerful SoC on paper. The new XBOX "Scorpio" 4K Ultra HD game console will feature a custom-design SoC by AMD, which will combine not just a GPU based on the "Polaris" architecture, but also a CPU based on the "Zen" microarchitecture. This is significant because it sees a departure from using 8 smaller "Jaguar" CPU cores, and upshifts to stronger "Zen" ones. The chip could be built on the 14 nm process.
The SoC powering the XBOX Scorpio could feature a CPU component with eight "Zen" CPU cores, with SMT enabling 16 logical CPUs, and a "Polaris" GPU with 6 TFLOP/s of compute power. The combined compute power is expected to be close to 10 TFLOP/s. The Radeon RX 480, for instance features 5.84 TFLOP/s of power at its given clock speed. The CPU and GPU will likely share a common memory interface, belting out a memory bandwidth of 320 GB/s. The silicon muscle of this console should power 4K Ultra HD, 1080p @ 60 Hz HDR, and "good VR" solutions such as the Oculus Rift and HTC Vive. Games for the console could leverage DirectX 12.
Source:
TweakTown
The SoC powering the XBOX Scorpio could feature a CPU component with eight "Zen" CPU cores, with SMT enabling 16 logical CPUs, and a "Polaris" GPU with 6 TFLOP/s of compute power. The combined compute power is expected to be close to 10 TFLOP/s. The Radeon RX 480, for instance features 5.84 TFLOP/s of power at its given clock speed. The CPU and GPU will likely share a common memory interface, belting out a memory bandwidth of 320 GB/s. The silicon muscle of this console should power 4K Ultra HD, 1080p @ 60 Hz HDR, and "good VR" solutions such as the Oculus Rift and HTC Vive. Games for the console could leverage DirectX 12.
78 Comments on Microsoft XBOX Scorpio SoC Powered by "Polaris" and "Zen"
HDR coming soon. ;)
I would hazard to guess it is a cut down version of the rx480.
If the Jaguar core has graphics cores it could utilise them also.
Edit: I for one cannot await HDR. Waiting for a good 3.440 x 1.440 panel that does 100+ fps and supports HDR myself. This will also be the time when I upgrade the GPU.
I hate to think how much they are going to sting us for the Scorpio when it comes out.
Better? Yes. Comparable to computers released at the same time? Don't know but, I suspect not.
That leaves ~4 TFLOPS for 8 Zen cores, so... ~500 GFLOPS per one Zen core in a console APU? As much as I would like that to happen (healthy competition should bring prices down for everyone), I find it hard to believe.
Note that Zen is just speculation, not something MS or AMD has announced. (although reasonable one)
PS
I love how console upgrades are now backward compatible. Thank you Sony, Microsoft and, of course, AMD. :)
If you're just using that 22" Samsung though, yea, no difference at all..lol
- How much throughput do I get with any particular serial workload for whatever I'm doing (single core/execution unit performance, what can I do in lock-step)?
- How many cores do I have (how many times can I do the things above in a purely parallel manor)?
- What kind of latency am I introducing by using multiple cores?
I don't really care about what the full aggregate compute power is because in this day and age, that doesn't tend to be the bottleneck, your serial workloads tend to limit how parallel your applications can be. The things you have to do one after another on a single core to ensure that things happen properly is always what slows applications down from this standpoint and we can only use multiple when we're not adding so much overhead from using multiple cores that you end up getting nowhere (or even degrade performance.)That's my rant about how FLOPS are a terrible gauge of aggregate performance and (in my opinion,) means very little. You could have two very different machines with the same "FLOPS" capability but, with very real differences in performance when it comes to real world applications. Aside from that, it's not like I spend my entire day doing floating point operations, integer operations are kind of an important thing too but, once again per core or EU.
</rant>You can thank Intel too, x86 and DirectX are what makes it backwards compatible. It's a PC, didn't you know this? :pIt matters if the person is thinking about video playback and not gaming. If the iGPU is going to be as powerful as something like my 390, I doubt it will play games well at 4k by itself but, probably would have aboslutely no problem playing back 4k content or even playing some select games in 4k that might not be as graphically demanding as others.
For example, on the Xbox 360 a lot of games ran at 720P and upscaled to 1080p but, there were some games like Geometry Wars 2, which was an arcade game, which was simple enough where 1080p was more than realistic. I wouldn't be surprised if this new Xbox worked the same way, where if games were not graphically demanding enough to overload the GPU, that they could be run at a higher resolution. To me, that makes sense. It would be like me grabbing an old game (that somehow supported 4k,) and running it at that resolution. The game isn't demanding, so running it at 4k might be realistic.
Intel can't make them (suck on GPU side, deadly for consoles), nVidia can't make them (suck on CPU side) and if you buy separate GPU and CPU it's too expensive for a viable console.That's what FLOPS are. That's what GPUs do most of the time. It's pretty objective measure, although it doesn't necessarily translate 1:1 into gaming performance.
And back to the topic, Last time I've checked i7 965 had 0,07 TFlops.
Seems like consoles will be nothing more but custom designed PCs in the near future. If that means PC gamers won't miss out on exclusives anymore (hopefully Nintendo's as well) I'm all for it.
IBM willingness to do PC and Jobs' greediness is the reason why x86 is a thing.
Oh, and IBM would not have picked up lolwhatwhohaveheardaboutthiscompany Intel, if not AMD (that's why both co-existed in x86 space in the first place).
Instruction set doesn't matter at all, it's about sticking with one architecture,, whatever it is.
ARM isn't quite there yet (good luck, Nintendo, to me your decisions look idiotic) and we have no other major actively developed CPU architectures out there. CELL was a major disaster for Sony, weirdo CPU (1 normal core, 8 vector cores, go multi-platform for it) for R&D of which Sony alone spent 4 billion $. Now they pay like a hundred or two millions for new "semi-custom", when (if) they need it.Floating Point Operations Per Second (which GPUs can crunch plenty, unlike CPUs), which "per execution unit" did you read between which lines?That's Microsoft's view, not Sony's.