Intel's Architecture Day 2018: The Future of Core, Intel GPUs, 10nm, and Hybrid x86
by Dr. Ian Cutress on December 12, 2018 9:00 AM EST- Posted in
- CPUs
- Memory
- Intel
- GPUs
- DRAM
- Architecture
- Microarchitecture
- Xe
Intel’s First Fovoros and First Hybrid x86 CPU: Core plus Atom in 7 W on 10 nm
Perhaps someone will correct me, but I can’t ever remember a time when Intel has put multiple x86 cores of different configurations on the same bit of silicon (ed: Intel Edison). Ever since Arm starting doing it with its big.Little designs in smartphones, a perennial question was if Intel was going to do something similar, either with big and small Atom cores, or by moving a high-performance Core into the mix. When Intel left the smartphone and tablet market, we assumed the idea was dead. But, like a reanimated zombie, it has risen from the grave. Enter Intel’s Hybrid x86 CPU.
This tiny 12x12 package is built using Intel’s Fovoros technology, using a 22FFL IO chip as the active interposer connected with TSVs to a 10nm die that contains both a single Sunny Cove core and four Atom (Tremont?) cores. This tiny chip is smaller than a dime, and is designed to have a 2 mW standby power. It would appear that this chip is destined for mobile devices.
Here’s the manufacturing diagram, showing the idea that POP memory is placed over the Fovoros design to give the final product. Very much like a mobile chip.
The demo system that Intel had on display looked similar to the previous Sunny Cove design, however this heatsink was smaller and it had a few different connectors. We were told that this chip will support PCIe for M.2 as well as UFS, both of which are found in mobile. There also looked like a couple of SIM card connectors on this motherboard.
The key part of this discussion however is this block diagram that was on one of the Intel slides. Here we see a single ‘Big CPU’ with 0.5 MB of private medium level cache, four ‘Small CPU’s with a shared 1.5 MB L2 cache, an uncore that has 4MB of last level cache, a quad-channel memory controller (4x16-bit) with support for LPDDR4, a 64 EU design with Gen11 graphics, the Gen 11.5 display controller, a new IPU, MIPI support with DisplayPort 1.4, and all of this in a tiny package.
Seriously though, this has the potential to be a large revenue stream for Intel. They’ve made this chip, which allows the cores to enter C6 sleep states when not in use, that has a die size smaller than 12x12mm (144 mm2), and target the sub-7W fanless device market. That’s with a big Core, four Atom cores, and a GT2 64 EU design.
Intel actually says that the reason why this product came about is because a customer asked for a product of about this performance but with a 2 mW standby power state. In order to do this, Intel created and enhanced a number of technologies inside the company. The final product is apparently ideal for the customer, however the chip will also be made available for other OEMs.
In our Q&A session with the senior members of Intel, it was clear that this technology is still in its infancy, and Intel now has a new toy to play with. Jim Keller stated that internally they are trying lots of new things with this technology to see what works and what would make a good product, so we should be seeing more Foveros designs through 2019 and 2020.
Facebook
Twitter
RSS
148 Comments
View All Comments
zodiacfml - Thursday, December 13, 2018 - link
YESRaqia - Thursday, December 13, 2018 - link
For ultra-mobile, not only are battery/power/heat issues but supply is one as well due to Intel being locked down to their own manufacturing division. On top of that, they have a lock on x86 by not licensing to any competitors but AMD, who despite competitive stretches inevitably stumbles (either due to themselves or Intel's non-engineering financial efforts) and leaves the industry with dry spells of performance improvements. Intel's gross margins on their chips remain >60% as a result whereas ARM SoCs even after licensing is closer to 20-30%.Raqia - Thursday, December 13, 2018 - link
Keller declared that the technology is in its infancy, and feature wise the 2019 version of the Atom simply won't be competitive with leading ARM SoCs like the 8cx. The slowness you refer to only occurs when running native 32 bit x86 code on the WOW emulation layer, but the value of this feature is mostly in the compatibility being there at all. If performance and compatibility of legacy code matters to you then certainly Windows on Arm isn't suitable. However, it will matter even less now with the new native compilation tools and ports of important sub platforms like Chromium.29a - Thursday, December 13, 2018 - link
"Windows on ARM is horribly slow and therefore shitty."Sounds a lot like Windows on Atom.
MonkeyPaw - Saturday, December 15, 2018 - link
I’m betting Apple wanted one for MacBook Air, or maybe MS for Surface Go. It would be the right amount of performance for both devices, an both companies would have the clout to get it done. I’d lean toward Apple because the GPU is pretty big.Kevin G - Wednesday, December 12, 2018 - link
Typo:"a physical address space up to 52 bits. This means, according to Intel, that the server processors could theoretically support 4 TB of memory per socket."
That should be petabytes instead of terabytes. The limit is for an entire system, not per socket as additional sockets will not grant any additional capacity.
gamerk2 - Thursday, December 13, 2018 - link
NUMA systems could potentially be per-socket rather then OS wide.HStewart - Wednesday, December 12, 2018 - link
It sounds like Intel has been working on increasing performance in two ways1. 7nm change for the future - because of limitations found with 10nm
2. 10nm enhance for corrections for performance of issues with Cannon Lake
But most importantly, architexture improvements like faster single thread execution and new instructions and multi-core improvements will in long term significantly improve performance
ishould - Wednesday, December 12, 2018 - link
Forgive me if I take 2 metric tons of salt with any roadmaps Intel provides these days. They haven't exactly had the most accurate timelines as of late (past four years)HStewart - Wednesday, December 12, 2018 - link
It appears they realize that and coming out with document to indicated they have made corrections - this is better than not knowing what they are planning - or as some AMD Fans would like to believe that they lost the battle.