Cascade Lake
2nd Generation Intel® Xeon® Scalable Processors with Intel® C620 Series Chipsets (Purley Refresh)
2nd Generation Intel® Xeon® Scalable Processors, formerly Cascade Lake, with Intel® C620 Series Chipsets (Purley refresh), features built-in Intel® Deep Learning Boost and delivers high-performance inference and vision for AI workloads. It consolidates diverse IoT workloads, handles massive datasets and enables near-real-time transactions. Now you can get even better built-in deep learning capabilities, speed deployment, and lower total cost of ownership (TCO) with CPU-optimized software toolkits and frameworks such as Intel® Distribution of OpenVINO™ Toolkit.
Key Features
Intel® Deep Learning Boost
Accelerates AI/deep learning/vision workloads up to 14X 1 the inference throughput performance over previous generation processors.
Intel® Optane™ DC persistent memory
Speed workloads and time to insight with this new revolutionary memory product for affordable, persistent, and large memory.
Integrated Intel® QuickAssist Technology (Intel® QAT)
Data compression and cryptography acceleration, frees the host processor, and enhances data transport and protection across server, storage, network, and VM migration. Integrated in the chipset.
Intel® Resource Director Technology for Determinism
Extend Quality of Service (QoS) with memory bandwidth allocation.
Enhanced Security
Hardware mitigations for side-channel exploits help protect systems and data by hardening the platform against any malicious attacks.
Extended Availability of Support
15-year product availability and 10-year use-case reliability helps protect your investment.
Key Specifications
- Up to 28 CPU cores
- Up to 3.8 GHz non-AVX base CPU frequency
- Multi-socket support (2, 4, 8 CPU)
- Up to 3 UPI channels per CPU
- 6 channels DDR4 per CPU with max 2933MT/s speed
- 1 TB to 4.5 TB memory capacity per CPU
- Integrated Intel® Ethernet Connection X722
- 48 lanes of PCIe 3.0 per CPU
- Supports PCIe*, USB, SATA* and connects to Ethernet, SSD and FPGA peripherals
Highest Performance
Balanced Energy Efficiency
Extended Reliability & Memory
Specialized
Chipsets
Chipset | 10Gb/1Gb Ethernet Ports |
TDP (W) | PCIe* Uplink | Intel® QuickAssist Technology |
IoT Options Available |
Ordering Code |
---|---|---|---|---|---|---|
Intel® C629 Chipset | 4/4 | 28.6 | X16 | Yes | - | EY82C629 |
Intel® C628 Chipset | 4/4 | 26.3 | x16 | Yes | - | EY82C628 |
Intel® C627 Chipset | 4/4 | 28.6 | x16 | Yes | - | EY82C627 |
Intel® C626 Chipset | 4/4 | 23 | x16 | Yes | - | EY82C626 |
Intel® C625 Chipset | 4/4 | 21 | x16 | Yes | - | EY82C625 |
Intel® C624 Chipset | 4/4 | 19 | x16 | - | Yes | EY82C624 |
Intel® C622 Chipset | 2/4 | 17 | x8 | - | Yes | EY82C622 |
Intel® C621 Chipset | 0/4 | 15 | x1 | - | Yes | EY82C621 |
Supported Software
OS Type | Operating System 2 (Targeted for Support) | Support 3 | Distribution | BIOS |
---|---|---|---|---|
Linux | Red Hat* Enterprise Linux 7.5 | Red Hat | American Megatrends Inc Insyde Software Phoenix Technologies BYOSOFT |
|
SUSE* Linux Enterprise Server 12 SP4, 15 | SUSE, Open Source | SUSE | ||
Ubuntu* 18.04 LTS | Canonical, Open Source | Canonical | ||
Yocto* Linux v4.19.8 | Intel, Open Source | Yocto Project* | ||
FreeBSD 11.2 | Open Source Community | |||
Fedora* | Open Source Community | |||
CentOS* | Open Source Community | |||
Windows* | Microsoft Windows* Server 2016 Microsoft Windows* Server 2019 LTS Microsoft Windows* Server RS3, RS4, RS5 (Core/Nano) |
Intel, Microsoft | Microsoft | |
VMM | Linux KVM | Open Source Community | ||
VMware ESXi* 6.0 u3, 6.5 | VMware*, Open Source | |||
Microsoft Windows* Hyper-V | Microsoft | |||
Xen* 4.10, 4.11 | Open Source Community |
Software Tools
Intel® System Studio
Boost performance, power efficiency, and reliability for system and IoT device applications with this all-in-one development tool suite (Windows*, Linux, Android*, VxWorks*, QNX Neutrino RTOS*).
Intel® Distribution of OpenVINO™ Toolkit
Make your vision a reality on Intel® platforms—from smart cameras and video surveillance to robotics, transportation, and more (Windows*, Linux, CentOS*).
Intel® Data Analytics Acceleration Library
Boost big data analytics and machine learning performance with this easy-to-use library. (Windows*, Linux, macOS*).
Intel® Distribution of Python*
Supercharge Python* applications and speed up core computational packages with this performance-oriented distribution (Windows*, Linux, macOS*).
Embedded & IoT Optimized Applications
Smart Cities
Whether densely populated or remote, AI applications with Intel® Deep Learning Boost support faster, more accurate security and surveillance even in crowded, complex urban environments
Healthcare
Object detection and segmentation identify and compare relevant patterns and other imaging data faster and more accurately, which speeds and improves diagnoses, delivering better outcomes for more patients, and reduced costs for hospitals
Industrial & Manufacturing
Intel® Deep Learning Boost brings the performance and capabilities that accelerate industrial IoT and manufacturing to advance AI, increase performance, use machine vision for defect detection and quality inspection, and consolidate workloads
Design Resources
Design-In Tools Store
Accelerate your design process with tools that support our latest platforms. All tools are available for purchase, plus Intel offers a limited selection of embedded development tools for loan at no cost to developers who meet the loan program’s criteria.
Free Design Review Services
Speed your design cycle with Intel’s free-of-charge schematic and layout reviews.
Free Layout Review Services
Optimize system performance and product design with our comprehensive testing services.
IoT Developer Resources
Leverage Intel’s tools, kits, and solutions to speed your time to market.
Building Your Application
Optimization
Reference Links
Product and Performance Information
1x inference throughput improvement on Intel® Xeon® Platinum 8180 processor (July 2017) baseline: Tested by Intel as of July 11th 2017: Platform: 2S Intel® Xeon® Platinum 8180 CPU @ 2.50GHz (28 cores), HT disabled, turbo disabled, scaling governor set to “performance” via intel_pstate driver, 384GB DDR4-2666 ECC RAM. CentOS Linux* release 7.3.1611 (Core), Linux kernel 3.10.0-514.10.2.el7.x86_64. SSD: Intel® SSD DC S3700 Series (800GB, 2.5in SATA 6Gb/s, 25nm, MLC). Performance measured with: Environment variables: KMP_ AFFINITY=’granularity=fine, compact‘, OMP_NUM_THREADS=56, CPU Freq set with cpupower frequency-set -d 2.5G -u 3.8G -g performance. Caffe: (http://github.com/intel/caffe/), revision f96b759f71b2281835f690af267158b82b150b5c. Inference measured with “caffe time --forward_only” command, training measured with “caffe time” command. For “ConvNet” topologies, synthetic dataset was used. For other topologies, data was stored on local storage and cached in memory before training. Topology specs from https://github.com/intel/caffe/tree/master/models/intel_optimized_models (ResNet-50), and https://github.com/soumith/convnet-benchmarks/tree/master/caffe/imagenet_winners (ConvNet benchmarks; files were updated to use newer Caffe prototxt format but are functionally equivalent). Intel® C++ Compiler ver. 17.0.2 20170213, Intel® Math Kernel Library (Intel® MKL) small libraries version 2018.0.20170425. Caffe run with “numactl -l“.
14x inference throughput improvement on Intel® Xeon® Platinum 8280 processor with Intel® Deep Learning Boost (Intel® DL Boost): Tested by Intel as of 2/20/2019. 2 socket Intel® Xeon® Platinum 8280 processor, 28 cores HT On Turbo ON Total Memory 384 GB (12 slots/ 32GB/ 2933 MHz), BIOS: SE5C620.86B.0D.01.0271.120720180605 (ucode: 0x200004d), Ubuntu 18.04.1 LTS, kernel 4.15.0-45-generic, SSD 1x sda INTEL SSDSC2BA80 SSD 745.2GB, nvme1n1 INTEL SSDPE2KX040T7 SSD 3.7TB, Deep Learning Framework: Intel® Optimization for Caffe* version: 1.1.3 (commit hash: 7010334f159da247db3fe3a9d96a3116ca06b09a), ICC version 18.0.1, MKL DNN version: v0.17 (commit hash: 830a10059a018cd2634d94195140cf2d8790a75a, model https://github.com/intel/caffe/blob/master/models/intel_optimized_models/int8/resnet50_int8_full_conv.prototxt, BS=64, syntheticData, 4 instance/2 socket, Datatype: INT8 vs. Tested by Intel as of July 11, 2017: 2S Intel® Xeon® Platinum 8180 CPU @ 2.50GHz (28 cores), HT disabled, turbo disabled, scaling governor set to “performance” via intel_pstate driver, 384GB DDR4-2666 ECC RAM. CentOS Linux* release 7.3.1611 (Core), Linux kernel 3.10.0-514.10.2.el7.x86_64. SSD: Intel® SSD DC S3700 Series (800GB, 2.5in SATA 6Gb/s, 25nm, MLC). Performance measured with: Environment variables: KMP_AFFINITY=’granularity=fine, compact‘, OMP_NUM_THREADS=56, CPU Freq set with cpupower frequency-set -d 2.5G -u 3.8G -g performance. Caffe: (https://github.com/intel/caffe/), revision f96b759f71b2281835f690af267158b82b150b5c. Inference measured with “caffe time --forward_only” command, training measured with “caffe time” command. For “ConvNet” topologies, synthetic dataset was used. For other topologies, data was stored on local storage and cached in memory before training. Topology specs from https://github.com/intel/caffe/tree/master/models/intel_optimized_models/resnext_50, Intel® C++ Compiler ver. 17.0.2 20170213, Intel® MKL small libraries version 2018.0.20170425. Caffe run with “numactl -l“.
This is the OS list that is tested internally and does NOT reflect the OS vendor support for these exact release versions. Please contact respective OS vendor(s) for the release version numbers and support. Several software patches will be upstreamed and will be picked up over time. These will be required to enhance platform support.
Intel only provides support for our tools, patches and utilities on the OS. Actual OS support should come from the OS Vendor.