Thermal Management on Stacked Silicon

With a standard processor design, there is a single piece of silicon doing all the work and generating the heat – it’s bonded to the package (which doesn’t do any work) and then depending on the implementation, there’s some adhesive to either a cooler or a headspreader then a cooler. When moving to a stacked chiplet design, it gets a bit more complicated.

Having two bits of silicon that ‘do work’, even if one is the heavy compute die and the other is an active interposer taking care of USB and audio and things, does mean that there’s a thermal gradient between the silicon, and depending on the bonding, potential for thermal hotspots and build-up. Lakefield makes it even more complex, by having an additional DRAM package placed on top but not directly bonded.

We can take each of these issues independently. For the case of die-on-die interaction, there is a lot of research going into this area. Discussions and development about fluidic channels between two hot silicon dies have been going on for a decade or longer in academia, and Intel has mentioned it a number of times, especially when relating to a potential solution of its new die-to-die stacking technology.

They key here is hot dies, with thermal hotspots. As with a standard silicon design, ideally it is best to keep two high-powered areas separate, as it gives a number of benefits with power delivery, cooling, and signal integrity. With a stacked die, it is best to not have hotspots directly on top of each other, for similar reasons. Despite Intel using its leading edge 10+ process node for the compute die, the base die is using 22FFL, which is Intel’s low power implementation of its 14nm process. Not only that, but the base die is only dealing with IO, such as USB and PCIe 3.0, which is essentially fixed bandwidth and energy costs. What we have here is a high-powered die on top of a low powered die, and as such thermal issues between the two silicon die, especially in a low TDP device like Lakefield (7W TDP), are not an issue.

What is an issue is how the compute die gets rid of the heat. On the bottom it can do convection by being bonded to more silicon, but the top is ultimately blocked by that DRAM die. As you can see in the image above, there’s a big air gap between the two.

As part of the Lakefield design, Intel had to add in a number of design changes in order to make the thermals work. A lot of work can be done with the silicon design itself, such as matching up hotspots in the right area, using suitable thickness of metals in various layers, and rearranging the floorplan to reduce localized power density. Ultimately both increasing the thermal mass and the potential dissipation becomes high priorities.

Lakefield CPUs have a sustained power limit of 7 watts – this is defined in the specifications. Intel also has another limit, known as the turbo power limit. At Intel’s Architecture Day, the company stated that the turbo power limit was 27 watts, however in the recent product briefing, we were told is set at 9.5 W. Historically Intel will let its OEM partners (Samsung, Lenovo, Microsoft) choose its own values for these based on how well the design implements its cooling – passive vs active and heatsink mass and things like this. Intel also has another factor of turbo time, essentially a measure of how long the turbo power can be sustained for.

When we initially asked Intel for this value, they refused to tell us, stating that it is proprietary information. After I asked again after a group call on the product, I got the same answer, despite the fact that I informed the Lakefield team that Intel has historically given this information out. Later on, I found out through my European peers that in a separate briefing, they gave the value of 28 seconds, to which Intel emailed me this several hours afterwards. This value can also be set by OEMs.

Then I subsequently found one of Intel’s ISSCC slides.

This slide shows that a basic implementation would only allow sustained power for 2.5 seconds. Adding in an adhesive between the top die and the DRAM moves up to 12.4 seconds, and then improving the system cooling goes up to 20 seconds. The rest of the improvements work below the compute die: a sizeable improvement comes from increasing the die-to-die metal density, and then an optimized power floor plan which in total gives sustained power support for 150+ seconds.

Lakefield: Top Die to Bottom Die Hybrid CPUs: Sunny Cove and Tremont
POST A COMMENT

224 Comments

View All Comments

  • Santoval - Saturday, July 4, 2020 - link

    AVX2/512 is not the only bit that differentiates the Core series from Atom. If you only care about floating point performance then yes, that's their primary difference. AVX has nothing to do with integer code though. Reply
  • ProDigit - Friday, July 3, 2020 - link

    Not a single program, other than benchmarks use it.
    Very few use avx 256 even...
    I wonder why anyone would need it on a laptop, especially considering it's a server feat.
    Reply
  • Meteor2 - Friday, July 3, 2020 - link

    Not photo and video editing software? Reply
  • lefty2 - Friday, July 3, 2020 - link

    Tremont doesn't support any version of AVX. So, that would cause a huge swath of software to operate very slow. AVX is used a lot more universally than you think - practically all games use it. All multimedia applications, photo editing software, etc, etc. Reply
  • npz - Friday, July 3, 2020 - link

    You're wrong. AVX(2) is used almost universally among multimedia applications now. AVX512 is still rare mostly because of limited chip implementations, but AVX2 (the 256 bit version) is implemented everywhere in x86 performant applications. Nearly anything to do with video and you'll be using it Reply
  • dotjaz - Saturday, July 4, 2020 - link

    AVX512 is rare because it's pointless. For example, enabling AVX512 for x265 causes performance to drop below AVX level because of thermal throttling. Reply
  • npz - Saturday, July 4, 2020 - link

    That's just Intel"s implementation through, which also occurs using AVX2 but to a lesser degree. That's why intel bios have an AVX multiplier offset limit. AMD doesn't suffer from the same issue.

    But even with avx512 as long as you have say, no more than half the utilizing it, and can multitask to rotate or just use other cores fir
    Reply
  • npz - Saturday, July 4, 2020 - link

    ... fot other processes instead of purely just the avx512 program then you can still make some gains. Again it takes profiling and careful usage given the intel limitations. And yes, I saw the exact x265 test. They made gains per thread until a certain amount was used Reply
  • dotjaz - Saturday, July 4, 2020 - link

    Wrong, AVX2 is quite universal on any multimedia tasks. AVX512 is pointless. Reply
  • Samus - Saturday, July 4, 2020 - link

    Clearly a first-gen product. Give it time.

    For my sake I hope this isn't a dud because Intel clearly invested billions into the thing and I own a lot of Intel stock lol.
    Reply

Log in

Don't have an account? Sign up now