Hot on the heels of Intel's announcement of new mobile-focused CPUs integrating AMD Radeon graphics, we have our first glimpse at a real-world design using this new chip.
Posted on the infamous Chinese tech forum, Chiphell earlier today, this photo appears to be a small form factor PC design integrating the new Kaby Lake-G CPU and GPU solution.
Looking at the standard size components on the board like the Samsung M.2 SSD and the DDR4 SODIMM memory modules, we can start to get a better idea of the actual size of the Kaby Lake-G module.
Additionally, we get our first look at the type of power delivery infrastructure that devices with Kaby Lake-G are going to require. It's impressive how small the motherboard is taking into account all of the power phases needed to feed the CPU, GPU, and HBM 2 memory.
Looking back at the leaked NUC roadmap from September, the picture starts to become more clear. While the "Hades Canyon" NUCs on this roadmap threw us for a loop when we first saw it months ago, it's now clear that they are referencing the new Kaby Lake-G line of products. The plethora of IO options from the roadmap, including dual Gigabit Ethernet and 2 Thunderbolt 3 ports also seem to match closely with the leaked NUC photo above.
Using this information we also now have a better idea of the thermal and power requirements for Kaby Lake-G. The base "Hades Canyon" NUC is listed with a 65W processor, while the "Hades Canyon VR" is listed with as a 100W part. This means that devices retain the same levels of CPU performance from the existing Kaby Lake-H Quad Core mobile CPUs which clock in at 35W, plus roughly 30 or 65W of graphics performance.
These leaked 3DMark scores might give us an idea of the performance of the Hades Canyon VR NUC.
One thing is clear; Hades Canyon will be the highest power NUC Intel has ever produced, surpassing the 45W Skull Canyon. Considering the already unusual for a NUC footprint of Skull Canyon, I'm interested to see the final form of Hades Canyon as well as the performance it brings!
With what looks to be a first half 2018 release date on the roadmap, it seems likely that we could see this NUC or other similar devices being shown off at CES in January. Stay tuned for more continuing coverage of Intel's Kaby Lake-G and upcoming devices featuring it!
I wonder if this is why Tim
I wonder if this is why Tim Cook decided to publicly answer a user question regarding Apple’s plans for the Mac Mini a couple weeks ago. I’m sure i’m not the only one out there who looks at this new NUC and thinks, “that would make a splendid Hackintosh HTPC.”
I’m waiting to buy one so
I’m waiting to buy one so that I can install Qubes OS on it and run multiple emulated secure and fenced in unsecure panes. The Skull Canyon has a wicked overheating problem. Hopefully, Intel will fix that with this offering, and keep that vPro feature off of it. It’s bad enough they are running a separate MINIX OS to run their security subversive Intel ME on the i-series chips. It’s such a threat that even Google is scrambling to disable ME from their systems.
While I am surprised by the
While I am surprised by the number of power phases on that board, keep in mind that recent processors like this are truely SoC (or SoM) and that all that’s left to be done is power regulation. And they even tried to bring that on the chip/module.
This one could be a higher
This one could be a higher wattage part, possibly still under development. Looking at the power requirements, it may have the GPU performance of PlayStation4 or better. And you could be right about being a true SoC as it seems there isn’t any PCH or Southbridge chip on the board itself. Also, besides the M.2 slot, there isn’t any dedicated PCI Express expansion slots.
More than likely this NUC
More than likely this NUC will be more powerful than the original PS4 and Xbox1, they would be wise to launch this for around $399 as low budget gaming pc.
I expect the chip itself will
I expect the chip itself will cost more than that.
Yeah I would too, it can’t be
Yeah I would too, it can’t be a cheap solution.
Perhaps, but do note that
Perhaps, but do note that HBM2 is rather expensive.
On Anandtech’s podcast(1)
On Anandtech’s podcast(1) they estimated the custom Radeon GPU’s die size to be around that of Polaris 10 die size so this Intel NUC/EMIB MCM will eat into Microsoft’s and Sony’s PS4 business also Ryan Smith and Ian Cutress also speculated that that Intel’s CPU die size on the EMIB/MCM was similar to its 6 core Coffee Lake CPU.
This NUC could eat the Microsoft XBONE-X’s lunch and Redmond had better consider allowing XBONE-X’s users make use of the XBONE-X as a mini PC asme as this Intel NUC is a full PC. 8 of AMD’s Jag+ cores on the XBONE-X can not compete even with a 4 core Intel x86 variant so if this is a 6 core Intel CPU in that NUC with maybe some custom Polaris amount of Compute units on the Radeon die that is close to the XBONE-X’s CU count then folks will seriously go with Intel EMIB based CPU and Radeon/HBM2 mashup. Microsoft better turn the XBONE-X’s replacment into more of a full PC with the ability to run win32 apps and not just UWP apps, or allow the XBONE-X to run win32 applications.
I’d love to see the XBONE-X able to run graphics software in addition to games and even other applications and UWP apps are not very plentiful even now.
“The AnandTech Podcast, Episode 42: Intel with Radeon Graphics”
found my next pc 😀
found my next pc 😀
one word… STEAMBOX!!
one word… STEAMBOX!!
This is nice I’d rather go
This is nice I’d rather go with this than a laptop but Intel needs to get more Polaris CUs than 24.
I really think that AMD will have some realy nice Zen/Vega and HBM2 workstation grade full Interposer based APUs for portable workstations or even mini-desktop grade wirkstations Branded as Radeon Pro WX APUs and using AMD’s professional GPU drivers. It’s looks like AMD is happy enough letting Intel get AMD’s discrete die graphics inside of thin and light form factor laptops. So AMD will maybe be focusing on creating a workstation Zen cores, Vega die, and HBM2 full interposer based APU that can be sold for around the $1000-$1500 range because it will include a larger Vega die/pro branding and drivers and 8 or more Zen cores and 2/more stacks of HBM2.
Intel has so much market penetration in the thin and light with Nvidia’s Discrete GPU laptop market that AMD decided that this was a better way of getting into that market via suppling discrete radeon dies directly to Intel.
But AMD will not be partnering with Intel for any CPU/Vega/HBM2 workstation variants as AMD will be using Zen and its infinity fabric to create better workstation pro APU variants. That infinity Fabric IP supported in Zen and Vega will allow AMD to offer much tighter Zen CPU cores to Vega nCU Cache to Zen cores Cache coherency than that PCIe based Intel core to Polaris on that Intel MCM/EMIB solution.
Is this related to the
Is this related to the rumored Zeplin SoC?
The Zeppelin SOC is the
The Zeppelin SOC is the Zeppelin Die and on servers SOC means CPUs/processors that include a northbridge and a southbridge on the processor’s die and not necessarly any included Graphics.
So the Zeppelin die is that modular die that AMD uses across all of its Epyc/Ryzen/Threadripper SKUs. And that very same Zeppelin die is binned down to become the Summit Ridge platform Die with some server and other server grade on Zeppelin die functionality disabled. Summint Ridge is more of a platform code name for the consumer Ryzen binned Zeppelin dies that have Ryzen/socket AM4 support and the Threadripper Platform die is the same Zeppelin die binned down to support the X399 motherboard platform.
AMD has code names for CPU micro-archs(Zen) and GPU micro-archs(Vega), code names for socket/motherboard platforms and code names for dies(Zeppelin), Summit Ridge(binned Zeppelin Die) platform etc. Some codenames can become official marketing names and be used that way. Ryzen is not the name of a CPU micro-arch Ryzen is a brand name for a line of consumer SKUs that are made from the Binned Zeppelin die that have CPU cores that are based on the Zen CPU micro-arch.
AMD needs to publish a decoder ring code naming terminology glossery but the best place to find code name decoder rings is Xorg as they have good info for GPUs and CPUs related code names for the open source programming community.
A lot of systems/driver programming requires programmers to search for strings to identify what processor the system is using so that done with string searches where the processor’s code name is used to identify that processor along with ID numbers and such. Even platform names are included in header files/memory strings as that’s a way for compilers/software to query the OS about the harware and make use of the proper library functions for the PC’s CPU/MB chip-set/controllers: its processor ID name/number/stepping number/etc, its motherboard platform chipset related info, and and other system hardware attatched to a PC like peripherals, and such.