October 9th, 2024

Who are AMD, Intel's new manycore monster CPUs for?

Intel and AMD are launching high-core-count CPUs for server consolidation, but organizations should evaluate risks, costs, and disaster recovery capabilities before adoption, as hyperscale providers are better equipped for manycore systems.

Read original articleLink Icon
Who are AMD, Intel's new manycore monster CPUs for?

Intel and AMD are introducing high-core-count CPUs, such as Intel's 128-core Granite Rapids Xeon 6 and AMD's 192-core Turin Epyc, aimed at server consolidation. These processors promise to reduce rack space and power consumption significantly. However, the article cautions organizations against rushing to adopt this technology due to the increased risk associated with manycore systems. A single failure in a manycore server could lead to catastrophic consequences, and organizations may not be equipped to handle such risks. Additionally, the high memory requirements for these CPUs could lead to increased costs, particularly if utilization rates are low. The article emphasizes the importance of evaluating disaster recovery capabilities and software licensing implications before transitioning to these new systems. It suggests that hyperscale cloud providers like AWS and Google are better suited for these manycore machines due to their expertise in managing hardware risks and operationalizing such technology. Ultimately, while the advancements in CPU technology are impressive, organizations should carefully consider their readiness and the potential challenges before making significant investments.

- Intel and AMD are launching high-core-count CPUs for server consolidation.

- Manycore systems increase risk, as a single failure can have severe consequences.

- High memory requirements may lead to increased costs and low utilization rates.

- Organizations should assess disaster recovery and software licensing before adopting new technology.

- Hyperscale cloud providers are better positioned to manage manycore systems effectively.

Link Icon 13 comments
By @peutetre - 4 months
Video encoding. To get the best quality for bitrate you need to use a software encoder, and to get the best encode time you need to give it CPU resources.

It's been impressive how much SVT-AV1 has increased performance between releases. SVT-AV1 2.2 is a significant step up from 2.1:

https://www.phoronix.com/news/SVT-AV1-2.2-Released

https://gitlab.com/AOMediaCodec/SVT-AV1

By @noizejoy - 4 months
I’m fascinated by the idea of my future music making DAW computer having one of those manycore monsters.

That’s also because my favourite software synthesizers are increasingly modelling instruments, rather than sample based instruments. And that reduces the need for RAM and storage, but increases the thirst for CPU cycles.

By @mbrumlow - 4 months
Me?

I compile most of my OS and I would like it faster. I also like being able to compile and game at the same time. Or run many OSs in VMs at the same time.

By @teleforce - 4 months
Imagine one of this 128-core CPUs utilizing several TBs of RAM connected via CXL and several PBs of storage combine with real-time Linux, and I'm in software-defined-X (SD-x for SDN, SDR, etc) workstation wonderland [1], [2].

[1] Samsung Unveils CXL Memory Module Box: Up to 16 TB at 60 GB/s:

https://www.anandtech.com/show/21333/samsung-unveils-cxl-mem...

[2] Huawei unveils its OceanStor A800 AI-specific storage solution; announces 128TB high-capacity SSD

https://www.datacenterdynamics.com/en/news/huawei-unveils-it...

[3] Real-time Linux is officially part of the kernel:

https://news.ycombinator.com/item?id=41594862

By @m463 - 5 months

  make -j$(nproc)
By @physicsguy - 5 months
Supercomputers? I've done work on 192-core machines in the past doing MPI jobs.
By @credit_guy - 4 months
It looks to me Intel will cannibalize its older, 5th, generation Xeon processors. Here's a comparison of two 64 core processors, one is 5th generation, launched in Q4 '23 [1], and one is 6th generation, launched in Q2 '24 [2]. The first one as an MSRP of $12400 (not a mistake) and the second $2749. It is true that the 5th generation CPU has 128 threads, vs only 64 for the 6th generation, but is this worth a price premium of 350% ?

[1] https://www.intel.com/content/www/us/en/products/sku/237252/...

[2] https://www.intel.com/content/www/us/en/products/sku/240363/...

By @wyldfire - 4 months
It's tremendously difficult to keep 192 cores truly busy at the same time. Unless you have enormous caches and enormous memory throughput, that is.
By @bhouston - 4 months
Yeah, they are mostly for cloud providers. They will be running all of our docker containers and Kubernetes clusters and function as a service workflows as well as our databases.

It really should mean that cloud data centres should be able to greatly increase capacity without getting larger in terms of physical size. That is a huge net win for the cloud providers.

By @andrewstuart - 4 months
Me. The more CPU the better for all my systems.
By @bubblesnort - 4 months
I wonder how hard it is to kill a fork bomb with 192 cores running full blast.
By @bhaney - 4 months
Me
By @vlovich123 - 4 months
TLDR: As you’d expect hyperscalers.