Silicon, Glue, & NUMA Too

In the Ryzen family, AMD designed an 8 core silicon die known as a Zeppelin die. This consisted of two core complexes (CCX) of four cores each, with each CCX having access to 8 MB of L3 cache. The Zeppelin die had access to two DRAM channels, and was fixed with 16 PCIe lanes for add-in cards. With Threadripper, AMD has doubled up the silicon.

If you were to delid a Threadripper CPU, you would actually see four silicon dies, similar to what an EPYC processor would have, making Threadripper a Multi Core Module (MCM) design. Two of these are reinforcing spacers – empty silicon with no use other than to help distribute the weight of the cooler and assist in cooling. The other two dies (in opposite corners for thermal performance and routing) are basically the same Zeppelin dies as Ryzen, containing eight cores each and having access to two memory channels each. They communicate through Infinity Fabric, which AMD lists as 102 GB/s die-to-die bandwidth (full duplex bidirectional), along with 78ns to reach the near memory (DRAM connected to the same die) and 133ns to reach the far memory (DRAM on another die). We confirmed those numbers on DDR4-2400 memory, also achieving 65 ns and 108 ns respectively using DDR4-3200. 


Despite this AMD slide showing two silicon dies, there are four units of silicon in the package. Only two of the dies are active, so AMD has 'simplified' the diagram'

By comparison, EPYC lists die-to-die bandwidth as 42.6 GB/s at DDR4-2666. This is because EPYC runs fabric links to three dies internally and one die externally (on the next socket), which maximizes all the links available. The dies in Threadripper only have to communicate with one other die, so has more flexibility. To that extent, we’re under the impression that Threadripper is using two of these links at 10.4 GT/s using the following method:

  • Die to Die for EPYC is quoted as 42.6 GB/s at DDR4-2667
  • Die to Die for Threadripper is quoted as 102.2 GB/s at DDR4-3200
  • 42.6 GB/s * 2 links * 3200/2667 = 102.2 GB/s
  • 42.6 GB/s * 3 links * 3200/2667 at 8.0 GT/s = 115.8 GB/s (too high)
  • 42.6 GB/s * 3 links * 3200/2667 at 6.4 GT/s = 92.6 GB/s (too low)

This configuration for AMD is essentially what the industry calls a NUMA configuration: non-uniform memory access. If left as it, it means that code cannot rely on a regular (and low) latency between requesting something from DRAM and receiving it. This can be an issue for high-performance code, which is why some software is designed NUMA-aware, so that it can intelligently pin the memory it needs to the closest DRAM controller, lowering potential bandwidth but prioritizing latency.

NUMA is nothing new in the x86 space. Once CPUs began shipping with on-die memory controllers rather than using an off-die memory controller in the Northbridge, NUMA became an inherent part of multi-socket systems. In this respect AMD was the leader here right from the start, as they beat Intel to on-die memory controllers for x86 CPUs by years. So AMD has been working with NUMA for years, and similarly NUMA has been the state of affairs for Intel's multi-socket server systems for almost a decade.

What's new with Threadripper however is that NUMA has never been a consumer concern. MCM consumer CPUs have been few and far between, and we'd have to go all the way back to the Core 2 Quad family to find a CPU with cores on multiple dies, which was a design that predates on-die memory controllers for Intel. So with Threadripper, this is the very first time that consumers – even high-end consumers – have been exposed to NUMA.

But more importantly, consumer software has been similarly unexposed to NUMA, so almost no software is able to take its idiosyncrasies into account. The good news is that while NUMA changes the rules of the game a bit, it doesn't break software. NUMA-aware OSes do the heavy lifting here, helping unaware software by keeping threads and memory accesses together on the same NUMA node in order to ensure classic performance characteristics. The downside to this is that much like an overprotective parent, the OS is going discourage unaware software from using other NUMA nodes. Or in the case of Threadripper, discouraging applications from using the other die and its 8 cores.


At a hardware level, Threadripper is natively two NUMA nodes

In an ideal world, all software would be NUMA-aware, eliminating any concerns over the matter. From a practical perspective however, software is slow to change and it seems unlikely that NUMA-style CPUs are going to become common in the future. Furthermore NUMA can be tricky to program for, especially in the case of workloads/algorithms that inherently struggle with "far" cores and memory. So the quirks of NUMA are never going to completely go away, and instead AMD has taken it upon themselves to manage the matter.

AMD has implemented BIOS switches and software switches in order to better support and control the NUMAness of Threadripper. By default, Threadripper actually hides its NUMA architecture. AMD instead runs Threadripper in a UMA configuration: a uniform memory access system where memory is sent to any DRAM and the return is variable in latency (e.g. ~100ns averaging between 78ns and 133ns) but focusing for a high peak bandwidth. By presenting the CPU to the OS as a monolithic, single-domain design, memory bandwidth is maximized and all applications (NUMA-aware and not) see all 16 cores as part of the same CPU. So for applications that are not NUMA-aware – and consequently would have been discouraged by the OS in NUMA mode – this maximizes the number of cores/threads they can use and the memory bandwidth they can use.


All 32 threads are exposed as part of a single monolithic CPU

The drawback to UMA mode is that because it's hiding how Threadripper really works, it doesn't allow the OS and applications to make fully informed decisions for themselves, and consequently they may not make the best decisions. Latency-sensitive NUMA-unaware applications that fare poorly with high core/memory latencies can struggle here if they use cores and memory attached to the other die. Which is why AMD also allows Threadripper to be configured for NUMA mode, exposing its full design to the OS and resulting in separate NUMA domains for the two dies. This informs the OS to keep applications pinned to one die when possible as previously discussed, and this mode is vital for some software and some games, and we’ve tested it in this review.

Overall, using a multi-silicon design has positives and negatives. The negatives end up being variable memory latency, variable core-to-core latency, and often redundancy in on-die units that don’t need to be repeated. As a result, AMD uses 400mm2+ of silicon to achieve this, which can increase costs at the manufacturing level. By contrast, the positives are in silicon design and overall yeilds: being able to design a single piece of silicon and repeat it, rather than design several different floor plans which multiplies up the design costs, and having the (largely) fixed number of wafer defects spread out over many more smaller dies.

By contrast, Intel uses a single monolithic die for its Skylake-X processors: the LCC die up to 10-core and HCC die from 12-core up to 18-core. These use a rectangular grid of cores (3x4 and 5x4 respectively), with two of the segments reserved for the memory controllers. In order to communicate between the cores, Intel uses a networking mesh, which determines which direction the data needs to travel (up, down, left, right, or accepted into the core). We covered Intel’s MOdular Decoupled Crossbar (MoDe-X) methodology in our Skylake-X review, but the underlying concept is consistency. This mesh runs at 2.4 GHz nominally. Prior to Skylake-X, Intel implemented a ring topology, such that data would have to travel around the ring of cores to get to where it needed to go.

With reference to glue, or glue-logic, we’re referring to the fabric of each processor. For AMD that’s the Infinity Fabric, which has to travel within the silicon die or out to the other silicon die, and for Intel that’s the internal MoDe-X mesh. Elmer’s never looked so complicated.

Feeding the Beast and CPU Top Trumps AMD’s Solution to Dual Dies: Creator Mode and Game Mode
Comments Locked

347 Comments

View All Comments

  • Vorl - Thursday, August 10, 2017 - link

    the answer to both of you is that "this is a High end PC processor, not a workstation CPU, and not a server CPU. That was clearly covered at the start of the article.

    If you want raw number crunching info, there will be other sites that are going to have those reviews, and really, maybe anandtech will review it in that light since it really is such a powerful CPU in another review for server stuff.

    Also, there is a LOT of value in having a standardized set of tests. Even if a few tests here and there are no longer valuable like PDF opening, the same tests being used across the board are important for BENCH. you can't compare products if you aren't using the same tools.

    Unfortunately AMD is ahead of the curve currently with massive SMP being given to normal consumers now at a reasonable price. It will take a little time for dev's to catch up and really make use of this amazing CPU.

    With the processing power in a CPU like this imagine the game mechanics that can be created and used, For those of us that are more interested in making this a reasonably priced workstation/server build for VMs etc, cool for us, but that isn't where this is being marketed, and it's not really fair to jump all over the reviewer for it.
  • Zstream - Thursday, August 10, 2017 - link

    Utter rubbish. This CPU is designed for a workstation build. Some a product labeled Xeon is a workstation CPU, but this isn't?
  • mapesdhs - Friday, August 11, 2017 - link

    Yeah, TR doesn't really look like something that's massively aimed at gamers, it has too many capabilities and features which gamers wouldn't be interested in.
  • pm9819 - Friday, August 18, 2017 - link

    AMD themselves call it a consumer cpu. Is Intel paying them as well
  • Lolimaster - Friday, August 11, 2017 - link

    It's a HEDT/workstation, a year ago people called Workstation a dual Xeon 8 cores, which a sole 1950X replicates.

    Intel draws a line not supporting ECC, AMD supports ECC in all their main cpu's server or not all the way back to Athlon 64.

    16cores/32threads, ECC, 64 pci-e lanes, upgrade path to 32cores/64threads with zen3. Smells Workstation to me.

    Another thing is server cpu's which EPYC is, with features tailored to it, like a massive core count with low clock speeds to maximize efficiency and damn expensive mobos without any gamerish gizmo, just think to put on building without looking at net. TR can do a bit of that too, but optimized to an all around performance and budget friendly.
  • Ian Cutress - Thursday, August 10, 2017 - link

    Dan sums it up. Some of these tests are simply check boxes - is it adequate enough.

    Some people do say that an automated suite isn't the way to do things: unfortunately without spending over two months designing this script I wouldn't have time for nearly as much data or to test nearly as many CPUs. Automation is a key aspect to testing, and I've spent a good while making sure tests like our Chromium Compile can be process consistent across systems.

    There's always scope to add more tests (my scripts are modular now), if they can be repeatable and deterministic, but also easy to understand in how they are set up. Feel free to reach out via email if you have suggestions.
  • Johan Steyn - Thursday, August 10, 2017 - link

    Ian, I understand that you see them as checkboxes, but this is not a normal CPU John doe is going to buy. It has a very specific audience and I feel you are missing that audience badly. I guy that buys this to use for rendering or 3Dstudio Max, is not going to worry about games. Yes, it would be a great bonus to also be OK at it. Other sittes even did tests of running rendering as well as play games at the same time. TR shined like a star against Intel. This is actually something that might happen in real life. A guy could begin a render and then while waiting, decide to play a game.

    I would not buy TR to open pdf's, would I?
  • Ian Cutress - Thursday, August 10, 2017 - link

    No, but you open things like IDEs and Premiere. A PDF test is a gateway test in that regard with an abnormally large input. When a workstation is not crunching hard, it's being used to navigate through programs with perhaps the web and documents in tow where the UX is going to be indicative of something like PDF opening.
  • Lolimaster - Friday, August 11, 2017 - link

    Including useless benchs not only you waste target audience time, you too having to write and upload images from that useless benchs instead of making the article more interesting.

    How about a "the destroyer for HEDT/Workstion", a typical productivy load + some gaming, out of a sudden people will get TWICE the cpu resources, they can do things they couldn't before on the same machine.

    They could get a dual socket mobo with 2x10c Xeons paying the hefty premium with pathetic clock speeds if they wante to game a bit while doing work, TR fixed that, with mass consumer type of gaming performance while reducing the multicore costs by more than half (cores counts + ECC support without paying intel tax).
  • Lolimaster - Friday, August 11, 2017 - link

    And that audience few months ago was limited to do their productivity thing with 6-8 cores or 10 paying the huge intel tax, probably they couldn't game without hurting other things and had a 2 secondary PC for killing time.

    With TR and the massive 16 core count they can finally do all of that off a single PC or focus the entire powerhorse when they need (leaving things do work during their sleep).

Log in

Don't have an account? Sign up now