IBM unveils the world’s first quad-core AI accelerator chip

The chip can be scaled for commercial use in hybrid-cloud environments

IBM logo on a brown background

IBM has unveiled the world’s first quad-core artificial intelligence (AI) accelerator chip, built using seven-nanometer (7nm) MOSFET technology.

The company says it has optimized the novel chip for low-precision workloads with support for many AI models.

"In a new paper presented at the 2021 International Solid-State Circuits Virtual Conference (ISSCC), our team details the world’s first energy-efficient AI chip at the vanguard of low precision training and inference built with 7nm technology," said IBM researchers Ankur Agrawal and Kailash Gopalakrishnan. 

"Through its novel design, the AI hardware accelerator chip supports a variety of model types while achieving leading edge power efficiency on all of them."

AI accelerators are specialized hardware designed to enhance AI applications’ performance, including deep learning, machine learning, and neural networks. They use in-memory computing or low-precision arithmetic, resulting in faster execution of large and complex AI algorithms

IBM claims its new AI accelerator chip is the first to include an ultra-low precision hybrid 8-bit floating-point (HFP8) format for training deep learning models in a silicon technology node (7 nm EUV-based chip). The chip can also self-maximize its performance by slowing down during high-power computation phases, thanks to an integrated power-management feature.  

Related Resource

Six reasons to accelerate remote asset monitoring with AI

How to optimise resources, increase productivity, and grow profit margins with AI

Why you should accelerate remote access monitoring with AI - whitepaper from IBMDownload now

Furthermore, IBM said its AI chip “routinely achieved more than 80% utilization for training and more than 60% utilization for inference” as compared to mainstream GPU utilizations, which are typically below 30%.

According to IBM, the chip technology can be scaled and deployed commercially to support large-scale deep learning models in the cloud.

“Our new AI core and chip can be used for many new cloud to edge applications across multiple industries. For instance, they can be used for cloud training of large-scale deep learning models in vision, speech and natural language processing using 8-bit formats (vs. the 16- and 32-bit formats currently used in the industry),” said IBM.

“They can also be used for cloud inference applications, such as for speech to text AI services, text to speech AI services, NLP services, financial transaction fraud detection and broader deployment of AI models in financial services.”

Featured Resources

Modern governance: The how-to guide

Equipping organisations with the right tools for business resilience

Free Download

Cloud operational excellence

Everything you need to know about optimising your cloud operations

Watch now

A buyer’s guide to board management software

How the right software can improve your board’s performance

The real world business value of Oracle autonomous data warehouse

Lead with a 417% five-year ROI

Download now

Recommended

IBM to sell off Watson Health assets
Acquisition

IBM to sell off Watson Health assets

24 Jan 2022
IBM ramps up sustainability efforts with Envizi acquisition
Business strategy

IBM ramps up sustainability efforts with Envizi acquisition

12 Jan 2022
LG Electronics joins the IBM Quantum Network
big data

LG Electronics joins the IBM Quantum Network

10 Jan 2022
The many faces of artificial intelligence
Sponsored

The many faces of artificial intelligence

15 Dec 2021

Most Popular

How to move Microsoft's Windows 11 from a hard drive to an SSD
Microsoft Windows

How to move Microsoft's Windows 11 from a hard drive to an SSD

4 Jan 2022
Synology DiskStation DS2422+ review: A cube of great capacity
network attached storage (NAS)

Synology DiskStation DS2422+ review: A cube of great capacity

10 Jan 2022
Microsoft Exchange servers break thanks to 'Y2K22' bug
email delivery

Microsoft Exchange servers break thanks to 'Y2K22' bug

4 Jan 2022