Home / Component / CPU / Intel 10nm Ice Lake-SP Xeon CPUs begin to leak via benchmark databases

Intel 10nm Ice Lake-SP Xeon CPUs begin to leak via benchmark databases

It looks like Intel’s upcoming Ice Lake-SP Xeon processors aren’t too far away. Signs of life were spotted this week on the SiSoftware Sandra database, revealing the similarly spec’d Xeon Platinum 8352Y and 8352S. There is one key difference between the two chips, however, as the Y variant is for dual-socket configurations while the S is used in 4-socket systems.

While these two CPUs are intended for multi-socket systems, only one of them was tested in a dual-socket set-up. The Xeon Platinum 8352S was seemingly tested on its own in a single-socket system, while the 8352Y was tested in a dual-socket system. Both benchmark database entries were found by @momomo_us and list several key specs for both 10nm processors.

In terms of specifications, both SKUs come with 32 cores and 64 threads, a 2.2GHz base clock with up to 3.4GHz boost, 40MB of L2 cache, 48MB of L3 cache, and a rated TDP of 205W. Ice Lake-SP processors will also introduce support for 8-channel DDR4-3200 memory and PCIe 4.0.

Exactly how well these new Xeons will hold up against AMD’s EPYC processors remains to be seen. Last year, Intel claimed the upcoming 32-core Ice Lake-SP processors would outperform the 64-core EPYC 7742 in “life sciences and FSI workloads”. The two benchmarks we currently have show Intel’s new Xeons beating AMD’s EPYC 7002 CPU in “processor arithmetic” and “processor multi-media” tests, but they lose out to AMD’s 64-core EPYC chip.

Intel Ice Lake-SP Xeon Processors are expected to release later this year. AMD is also due to launch its 3rd Gen EPYC processors later this month.

KitGuru says: Intel has been slow to upgrade to 10nm but the wait seems to be nearly over. Now we just have to wait for the big Ice Lake-SP Xeon VS EPYC ‘Milan’ showdown in the months to come. 

Become a Patron!

Check Also

Nvidia reveals ARM-based Grace CPU aimed at AI and HPC applications

Last night during Jensen's GTC 2021 keynote, Nvidia announced a new ARM-based CPU. Known as "Grace", Nvidia's latest CPU utilises ARM core architecture and aims to deliver 10x more performance for systems training AI models and other high-performance computing workloads.