Graphcore is hiring again after its SoftBank acquisition — but what do the roles tell us about its future focus?

Graphcore
(Image credit: Graphcore)

  • Recruitment drive would increase Graphcore's headcount by 20%
  • Announcement marks first big moves since SoftBank acquisition
  • Roles available in silicon design, software, and infrastructure

Graphcore has announced plans for a fresh hiring drive just months after its landmark acquisition by SoftBank.

The British AI chip manufacturer announced it has 75 open positions spanning a range of business functions, including silicon design, data center infrastructure management, hardware engineering, and AI research.

“Whether your expertise is in silicon, systems, software, AI research or any of the other functions that we are expanding, this is a chance to work at the leading-edge of your field,” said CEO Nigel Toon.

Graphcore hiring spree

The recruitment campaign represents a 20% increase in Graphcore’s overall workforce, and will include positions at all of its global offices.

The company currently has sites in Bristol, Cambridge, and London in the UK, along with offices in Gdansk, Poland, and Hsinchu City, Taiwan.

Graphcore was acquired by SoftBank in July 2024 in a deal described by Toon as a “tremendous endorsement” for the company. As part of the deal, the company became a wholly-owned subsidiary of the Japanese investment bank.

While financial terms of the acquisition were undisclosed, it’s believed the deal was valued at around $600 million, representing a significant markdown on its 2020 valuation of around $2.5 billion.

Toon described the deal as a “tremendous endorsement” for the company and said the acquisition is a “great outcome”.

What does the move say about Graphcore future plans?

In its announcement, Graphcore said the move builds on a pledge to “continue to invest in the creation of high-skilled jobs spanning a range of disciplines” in the wake of its acquisition.

The open positions do point toward its potential future plans, however, particularly with regard to chip design. Of the 75 roles Graphcore is recruiting for, 10 of these alone are in Silicon engineering design, suggesting the company plans to ramp up development of its next generation chip range.

Graphcore currently boasts three generations of silicon, with its most recent - the Bow Intelligence Processing Unit (IPU) - released in 2022. The Bow IPU delivers up to 350 teraflops of AI compute, and boasts a 40% increase in performance and 16% power efficiency compared to the company’s previous generation IPU.

However, unlike other processors available on the market, Graphcore’s IPUs do not use High Bandwidth Memory (HBM) technology. Instead, they rely on internal SRAM.

Competitor chips do use HBM though, including Intel, AMD and Nvidia.

With a raft of positions available in its silicon design department, this begs the question of whether Graphcore could shake up its design to include HBM capabilities.

Similarly, advertised roles in its infrastructure solutions and software departments may also point toward a change of tack for Graphcore, perhaps with its focus on data center capabilities sharpening.

It wouldn’t be alone here, either. Nvidia offers its own software and infrastructure options for AI inferencing. This has been crucial in its rapid ascendance throughout the generative AI boom.

AMD has also focused heavily on its own data center segment in recent years to keep pace with Nvidia.

You might also like

TOPICS
News and Analysis Editor, ITPro

Ross Kelly is News & Analysis Editor at ITPro, responsible for leading the brand's news output and in-depth reporting on the latest stories from across the business technology landscape.

Read more
Nvidia CEO Jensen Huang
Nvidia is preparing for the post-GPU AI era as it is reportedly recruits ASIC engineers to fend off competition from Broadcom and Marvell
Arm
Why would Arm's owner want to buy a rival to AMD and Intel? I can think of one reason
broadcom
Apple's new BFF, Broadcom, reveals three hyperscalers want to deploy 1,000,000 GPUs or XPUs by 2027; something that will make Nvidia wince
Cerebras WSE-3
DeepSeek on steroids: Cerebras embraces controversial Chinese ChatGPT rival and promises 57x faster inference speeds
A mockup of the Intel LGA 1851 motherboard socket
Intel quietly adds Jaguar Shores to its Gaudi AI Accelerator roadmap as it seeks to compete more fiercely against AMD and Nvidia
Nvidia HQ
Nvidia calls DeepSeek an 'excellent AI advancement' and praises the Chinese AI app's ingenuity
Latest in Pro
Stress
Complexity of IT systems could be increasing security risks for businesses
Ai tech, businessman show virtual graphic Global Internet connect Chatgpt Chat with AI, Artificial Intelligence.
CEOs think they might lose their jobs if they can't deliver on AI
A person using a desktop computer.
The role of automation in achieving sustainability goals
Cyber-security
The definitive guide to credential collaboration
Eurocom Raptor X18
At $15,000, this massive 256GB RAM laptop makes Apple's MacBook Pro look affordable, tiny and very, very slow
Squarespace
Build a website for less with 10% off Squarespace subscriptions
Latest in News
Stress
Complexity of IT systems could be increasing security risks for businesses
Ai tech, businessman show virtual graphic Global Internet connect Chatgpt Chat with AI, Artificial Intelligence.
CEOs think they might lose their jobs if they can't deliver on AI
Tony Hawk's Pro Skater 3+4
From Ace of Spades to Them Bones, Tony Hawk's Pro Skater 3+4's soundtrack is already looking excellent
The Google Gemini logo against a black background.
I tried Gemini's new AI image generation tool - here are 5 ways to get the best art from Google's upcoming Flash 2.0 built-in image upgrade
An image of the Samsung Galaxy S25 Ultra from a hands-on event
Samsung Galaxy S26 Ultra could resurrect an intriguing camera feature
Eurocom Raptor X18
At $15,000, this massive 256GB RAM laptop makes Apple's MacBook Pro look affordable, tiny and very, very slow