Ahead of the annual Supercomputing 2018 conference next week, Intel is today announcing part of its upcoming Cascade Lake strategy. Following on from its server-focused Xeon Scalable Skylake family, Intel has already pre-announced that Cascade Lake-SP will form the next generation, with a focus on compute and security. Today’s announcement is for a product family to run alongside Cascade Lake-SP, called Cascade Lake-AP, or Cascade-AP for short. Cascade-AP is going to be aimed at ‘advanced performance’. In order to implement this new processor family, Intel is combining multiple chips in the same package.

Scaling up to 48-Cores Per CPU

Intel is still keeping a lot of details for Cascade-AP under wraps for now, but what we do know is some high level specifications: Cascade-AP processors will be up to 48 cores, possibly with hyperthreading, and is aimed at dual socket servers, for a total of 96 cores in a 2S system. As a result, a single Cascade-AP is essentially a 2S Xeon setup on a single chip.

Each Cascade-AP processor will have 12 DDR4 DRAM channels, although maximum capacity of memory has not been announced. Connection between the processors will be the standard UPI connection as seen on current Xeon Scalable processors, although connection speed was not specified. It would appear that Cascade-AP is not designed to scale beyond a 2S system.

Each Cascade-AP lake processor will offer up to 48 cores, and in order to do this Intel is using a multi-chip package (MCP). At this point we believe that Intel is using two silicon dies in the package, specifically two ‘XCC’ dies which should be 28-core by design but will be only enabled up to 24-core per silicon die. Intel did confirm that these dies are connected by UPI over the package, and not with Intel’s EMIB technology. This is unfortunate, given how Intel has promoted its EMIB technology as an alternative to interposers, but we have yet to see it in a product that connects two high-powered chips together, as was originally envisioned when the technology was first announced. It does mean however that it can use the same silicon as will be found in the standard Xeon Scalable processors.

No information was given about TDP per processor, or pricing, PCIe lanes, memory capacity/support, frequencies, or variants. We were told that it will be launched at the same time as the full Cascade Lake family ‘in the first part of 2019’ (to directly quote). Intel did give some base performance numbers, suggesting that a processor scores 3.4x higher than an EPYC 7601 on Linpack and 1.3x higher on Stream Triad. Deep Learning inference, a major workload for these new parts, was quoted as scoring 17x the base Skylake-SP launch performance, which would be 1.55x the original quoted Cascade-SP launch performance with the new VNNI instructions.

Given the extra power requirements and extra memory requirements, we fully expect Intel to be using a new socket for this processor. Rather than LGA4367, with 4367-pins, there have been recent leaks suggesting that an LGA5903 socket be where Cascade-AP will end up, although Intel has not confirmed this. Intel did state that Cascade-AP will be shipping for revenue in Q3 2018 to its high-profile customers as part of its early access program, however widespread availability will be ‘in the first part of 2019’.

More Cores Per Socket, But Potentially a Worse 4S System

With this new type of processor, even given Intel’s previous competitive analysis of MCP designs, Intel is pushing up the number of cores per socket, and cores per 2S system. The 2S server market is one of Intel’s biggest across the whole company, and the idea is to offer more value. It’s hard to state the impact that Cascade-AP will have on the product portfolio, but it comes to mind that Intel is trying to condense a 4P system into a 2P system, though it probably won’t perform as well.

Take, for example, a 24-core 4P system, which has 96 cores total. Using Intel’s 3 UPI links, the communication between the four processors allows each processor to talk to each other processor in only one hop. Each processor has six channels of memory, access to 48 PCIe lanes, potential access to OmniPath, etc.

In a Cascade-AP two socket system, it all depends on how the links between the sockets are formed.

We already know that within a single package, the two silicon dies will talk to each other over UPI through the package. This is lower power than a socket-to-socket UPI link, and might take advantage of lower latency as well. That is all well and good.

But between the sockets is where questions will lie. If Intel is only using one UPI link between the sockets, then only one die from each socket will be connected. This is essentially a straight chain connectivity, meaning that to talk to a silicon die on the other socket, it may require two hops.

If Intel is using two UPI links between the sockets, then we have a situation similar to a Xeon Gold-2UPI system where the processors are arranged in a square and there is no cross-connection. This creates a communication imbalance between processors and memory which might be hard to predict unless the software can enumerate and deal with it effectively.

The other option is if Intel is using four UPI links between the sockets – two from each silicon die, to one each of the other silicon dies on the other processor. This is identical to a 4P 3UPI system, except that the dual socket configuration makes this all the more complex. It would require at least two UPI links to cross over each other with the traces on the motherboard. This either requires a thick motherboard (adds cost) or a control repeater chip that can do it easier. But even then, we’re dealing with a non-uniform memory architecture in each socket and between sockets with different latencies between the hops. At least in a natural 4S system, the arrangement can ensure latencies are equal.

There’s also the question of PCIe lanes. Intel has not given a number, but we suspect it to be anywhere between 48 (24 per die) or 96 (48 per die) per socket. Which would be identical to a 4S system again.

Further to that, a question of frequencies. A current 24-core Xeon Platinum runs at 205W. Intel is not going to put the TDP at 410W when it puts two together – it will have to lower the frequency to hit something more suitable for the large socket. Add into that the complexity of supplying 48 cores with enough power to be competitive. With separate chips, each socket can have higher frequencies.

The only saving grace here would be actual physical volume. Some datacenters are all about compute density, and for some providers looking to deploy mass x86 CPUs in their offering, this could be a winner with them. Intel has repeatedly cited that Xeon-based inference as a major source of customers, hence the inclusion of features like VNNI on Cascade Lake to help drive that market.

Timeline for Cascade Lake and Cascade Lake-AP

Intel stated that they are shipping processors for revenue to select customers as part of its early adopter program. The new parts will be launched in the first part of 2019, with the full Cascade Lake family (SP and AP) being launched at the same time. We asked a series of pertinent questions after our briefing, however Intel was unwilling to answer anything additional to our presentation. Some further information (or demonstrations) might be seen at the Supercomputing 2018 conference happening next week.

POST A COMMENT

91 Comments

View All Comments

  • Spunjji - Tuesday, November 06, 2018 - link

    He is probably both trolling and thus, by definition, also being an idiot. Reply
  • Lord of the Bored - Monday, November 05, 2018 - link

    Yup. Disappointed that it wasn't called out in the article. "We are going to be almost 4x faster if our imaginary numbers prove accurate AND you kneecap the Epyc."

    It isn't a good look for Intel, but it does show they have some serious concerns about the competition.
    Reply
  • philehidiot - Tuesday, November 06, 2018 - link

    They also used binaries in the benchmarks specifically compiled for Intel's processors. So they basically raced against AMD. Having removed half the tyres from the AMD vehicle as well as making AMD race on a dirt track whilst Intel were on the tarmac.

    Seems legit.
    Reply
  • Zizy - Monday, November 05, 2018 - link

    I expected simpler "glue 2x 28C parts", which would lead to 16 memory channels and a 56C monster with pretty decent number of PCIe as well - not quite as high as on AMD side, but still decent.
    Why would Intel make such a puzzling chip instead - what is the benefit here? Do they feel AMD isn't THAT much of a threat so they can afford offering something intermediate or what?

    Supercomputing 2018 will be interesting :)
    Reply
  • sgeocla - Monday, November 05, 2018 - link

    It's not that they don't know how much of a threat AMD is, it's just that they need to keep juggling 28 die yields, power draw, sku stack and margins. Reply
  • ManuelDiego - Monday, November 05, 2018 - link

    I'm guessing the benefit is winning some time while they solve their 10nm woes and have a really scalable architecture that allows them to compete with Epyc. I mean compete in economic terms, they clearly still have an IPC lead, although a shrinking one, but they need a way to make CPUs with far more than 28 cores at a reasonable price. For the time being they will glue together two XCC cores to at least remain competitive agains the coming Zen 2 Epyc CPUs in terms of core counts, while they figure out how to also compete in TCO, or at least to remain reasonably competitive. Reply
  • GruenSein - Monday, November 05, 2018 - link

    I'd argue that the perceived lead in single threaded performance is largely based on clock rate and more aggressive power management. Apart from workloads which make use of AVX512 an such the clock-for-clock and core-for-core performance is quite similar in most tasks. Reply
  • FullmetalTitan - Wednesday, November 07, 2018 - link

    I think the salient comparison here was Zen+ against Intel 8th gen Core parts, and the IPC difference at iso-clock speeds was south of 10%, used to be near 20%. By the time Intel actually updates their u-arch for real, they will be at parity with AMD on IPC Reply
  • psychobriggsy - Monday, November 05, 2018 - link

    Power consumption. Intel is stuck on 14nm and they've already reached the reasonable per-socket TDP limit with the 28C parts. Now they are sticking two of them together, disabling SMT (note that in their benchmarks they also disabled SMT on the Epyc parts for no appreciable reason apart from the obvious), and disabling 8 cores. I guess they want to stay under 350W TDP, and rely on high-end cooling solutions. They have one benchmark for HPC that they look good in (although perf/W will be interesting versus Epyc 2) even though any serious HPC customer would be looking at dedicated HPC accelerators. Reply
  • Spunjji - Tuesday, November 06, 2018 - link

    This is what really gets me. Their biggest "win" is in an area where you'd be an absolute idiot to use a CPU for your primary processing requirements.

    What one can interpret from this is that they feel like they're up against a wall, big time.
    Reply

Log in

Don't have an account? Sign up now