On Tuesday, Intel finally launched its 3rd Gen Ice Lake-SP Xeon CPUs, the American company’s first server lineup to be made on the 10nm process node. While the Ice Lake-SP lineup will be competing against AMD’s EPYC Milan family, it aims at the server segment.
Starting off with Intel’s 3rd Gen Ice Lake-SP Xeon CPUs’ details. It will be based on the 10nm+ process node and utilize the Sunny Cove core architecture. Since 2019, the Intel Sunny Cove x86 architecture has been around, and its first appearance was made on Intel’s 10th Gen Ice Lake processors for the notebook segment.
Tiger Lake has been of Intel’s interest since it is based on the Willow Cove x86 architecture and uses the 10nm SuperFin process node.
According to a report by WCCFTECH, some of the major upgrades that Intel’s 10nm+ for Ice Lake-SP Xeon CPU will deliver are as follows:
- 2.7x density scaling vs 14nm
- Self-aligned Quad-Patterning
- Contact Over Active Gate
- Cobalt Interconnect (M0, M1)
- 1st Gen Foveros 3D Stacking
- 2nd Gen EMIB
Some of the major highlights of today’s Intel Xeon announcement include:
- New 3rd Gen Intel Xeon Scalable processors — combined with FPGAs, Intel’s portfolio of Intel® Optane™ persistent memory and storage, Ethernet adapters, and optimized software solutions — deliver performance and workload optimizations across high-performance computing (HPC), hybrid-cloud, intelligent edge applications, and networking.
- New 3rd Gen Intel Xeon Scalable processors feature a flexible architecture with integrated AI acceleration with Intel® DL Boost technology, plus advanced security capabilities to protect data and application code with Intel® Software Guard Extension (Intel® SGX) and Intel Crypto Acceleration.
- “New 3rd Gen Intel Xeon Scalable processors are rapidly ramping, having shipped more than 200,000 units for revenue in the first quarter of 2021 with broad industry adoption across all market segments, including the world’s top cloud service providers set to deploy services; more than 250 design wins within 50 unique OxM partners; more than 15 major telecom equipment manufacturers and communications service providers readying POCs and networking deployments, and over 20 HPC labs and HPC-as-a-service environments leveraging our latest Xeon Scalable processors.”
“Intel’s Ice Lake-SP will ship in two die configurations, XCC (Extreme Core Count) and HCC (High Core Count). The XCC SKUs will feature 16, 18, 28, 32, 36, 38, and up to 40 cores. The HCC SKUs will feature 8, 12, 16, 18, 20, 24, 26, and up to 28 cores. The TDPs will range from 105, 135, 150, 165, 185, 205, 220, 235, 250, and all the way up to 270W for the flagship SKU. The XCC variants with 32, 36, 38, and 40 cores will be configurated at around 205-270W TDPs.”
Intel’s Xeon Ice Lake-SP lineup is split between (from top to lowest) Platinum, Gold, and Silver SKUs. The Platinum 8380, which would feature a base clock of 2.30 GHz, 40 cores, and a 270W TDP.
Compared to AMD’s flagship, cores and base clocks are lower in the Intel product, however, the company’s architecture and process nodes are tuned for higher clock speeds so that might mark a small win for them.
AVX-512 instruction set handily beats AMD’s existing Rome CPUs but only when running applications that fully leverage AVX-512 instructions while the standard applications would not see the same benefit on Intel CPUs.
The Ice Lake-SP CPUs will be supported by the LGA 4189 socket (socket P+) used by its Whitley platform. 2-way CPU support will be provided by the Whitley platform features, which will be interconnected with a UPI (Ultra Path Interconnect).
The CPUs will be connected to the Intel C620A chipset through DMI & the chipset itself will feature up to “20 PCIe Gen 3 lanes, 10 USB 3.0 ports, and 14 SATA Gen 3 ports.” As for the CPUs, they will offer up to “8-channel memory support in DDR4-3200 (1 DPC) or DDR4-2933 (2 DPC) modes. The Ice Lake-SP processors will feature up to 64 PCIe Gen 4.0 lanes.”
- Built-in AI acceleration: The latest 3rd Gen Intel Xeon Scalable processors deliver the AI performance, productivity, and simplicity that enable customers to unlock more valuable insights from their data. As the only data center CPU with built-in AI acceleration, turnkey solutions, and extensive software optimizations, the new processors make it possible to infuse AI into every application from edge to network to cloud. “The latest hardware and software optimizations deliver 74% faster AI performance compared with the prior generation and provide up to 1.5 times higher performance across a broad mix of 20 popular AI workloads versus AMD EPYC 7763 and up to 1.3 times higher performance on a broad mix of 20 popular AI workloads versus Nvidia A100 GPU.”
- Built-in security: “With hundreds of research studies and production deployments, plus the ability to be continuously hardened over time, Intel SGX protects sensitive code and data with the smallest potential attack surface within the system.”
- Built-in crypto acceleration: Across a host of important cryptographic algorithms, Intel Crypto Acceleration delivers breakthrough performance. Encryption-intensive workloads received by businesses such as online retailers who process millions of customer transactions per day can use this to protect customer data without impacting user response times or overall system performance.
“The 2nd Gen Optane DC Persistent memory would be supported on both platforms (Barlow Pass), offering up to 3200 MT/s and 15% bandwidth improvement in a 15W DIMM. Whitley would be able to support up to 3200 MT/s and 6 TB capacity per socket.”
The lineup, which is quite confusing if compared to the AMD EPYC parts, starts with 8 cores and goes all the way up to 40 cores. You can see the table below to get a sense of what Intel will be offering in its 3rd Gen Xeon Ice Lake-SP stack:
Buy the Intel Core i7-11700K: https://amzn.to/3myVsAI