Emerald Rapids 5th Generation Xeon Processor Specifications Leaked, Supports DDR5-5600 Memory and Up to 64 Cores

We haven’t seen the Intel Sapphire Rapids series yet , and information about the next generation of Intel HEDT processors has already surfaced.

Yuuki_ans has given us a lot more info on the upcoming 5th Gen Xeon , codenamed Emerald Rapids . These processors will use the same Eagle Stream platform , but with minor changes such as support for faster DDR5 memory.

Emerald Rapids

The Intel Emerald Rapids processors will be based on the Intel 7 process . Since this is the second iteration of the same node, expect better performance and increased efficiency. This line will ship with a maximum of 64 cores , which was 56 cores for the Sapphire Rapids. This is comparable to Milan in terms of core count, however AMD processors in Bergamo and Genoa will offer twice as many cores and go on sale sometime in 2023 .

The Raptor Cove -based Emerald Rapid will deliver 5-10% IPC gains over the previous generation. These processors will have TDPs ranging from 125W to 350W .

Faster memory and PCIe lanes

The transfer rate of 1DPC (1 DIMM per channel) will increase from 4800 MT/s (Sapphire Rapids) to 5600 MT/s . If you choose a 2DPC configuration (2 DIMMs per channel), the expected transfer rate is about 4800 MT/s .

The leaked specs say these processors will have 80 PCIe Gen 5.0 lanes . The new platform will support x16 , x8 , x4 and x2 (Gen4) forking along with shared virtual memory and scalable I/O virtualization .

The UPI or Ultra Patch Interconnect is rated at 20 GT/s , which is 25% more than the last generation’s 16 GT/s .

release date

While we have an exact release date, don’t expect these processors to arrive until late 2023 . Also, we’ll be spoiled with AMD offerings by then, so these processors may not take much market share. Are you excited about Intel Emerald Rapids processors? Team blue considers them a significant improvement over the latest generation (which has yet to be launched).

Emerald works great, high confidence in where we see what comes out.

~Intel