Intel reveals 3rd-gen 'Cooper Lake' Xeon Scalable processors

Chipmaker shows continued interest in AI workloads with bfloat16-based CPU

Intel has introduced the third-generation of its Xeon scalable processor, a range of chips designed for data centres and artificial intelligence workloads.

It's the industry's first mainstream server processor with built-in bfloat16 (brain floating point) support, according to Intel. The built-in bfloat16 is a compact numeric format which boosts the AI performance of its CPU, which makes AI inference and training more widely deployable on general-purpose CPUs.

It follows on from the 2nd-gen Xeon Scalable Processor, which was announced in February, designed to take on AMD's Epyc Rome family of data centre chips which focus on the 5G market.

The 3rd-gen, however, focuses on AI-based applications such as image classification, recommendation engines, speech recognition and language modelling.

By 2021, 75% of commercial enterprise apps will use AI, according to IDC. Along with analytics, its opening opportunities across a broad range of industries, including finance, healthcare, industrial, telecom and transportation.

"The ability to rapidly deploy AI and data analytics is essential for today's businesses," said Lisa Spelman, Intel's corporate VP and GM of Xeon and Memory Group.

"We remain committed to enhancing built-in AI acceleration and software optimisations within the processor that powers the world's data centre and edge solutions, as well as delivering an unmatched silicon foundation to unleash insight from data."

The 3rd Gen Xeon Scalable processor, which were codenamed 'Cooper Lake", is designed for deep learning, virtual machine density, in-memory database, mission-critical applications and analytics-intensive workloads.

In addition to the Xeon processor, which is available from today, the chipmaker is also launching Intel Optane persistent memory 200 series, with general OEM systems expected to be available later in the year. The Intel Optane persistent memory 200 series can provide up to 4.5 terabits of memory per socket to manage data-intensive workloads, according to Intel. 

Featured Resources

Choosing a collaboration platform

Eight questions every IT leader should ask

Download now

Performance benchmark: PostgreSQL/ MongoDB

Helping developers choose a database

Download now

Customer service vs. customer experience

Three-step guide to modern customer experience

Download now

Taking a proactive approach to cyber security

A complete guide to penetration testing

Download now

Most Popular

Microsoft is submerging servers in boiling liquid to prevent Teams outages
data centres

Microsoft is submerging servers in boiling liquid to prevent Teams outages

7 Apr 2021
How to find RAM speed, size and type
Laptops

How to find RAM speed, size and type

8 Apr 2021
UK exploring plans to launch its own digital currency
digital currency

UK exploring plans to launch its own digital currency

19 Apr 2021