Credit: IM Imagery/shutterstock. · Verdict · IM Imagery/shutterstock. GlobalData
The collaboration will also support the development of locally designed semiconductor products.
Malaysia has entered into a partnership with Arm Holdings, committing to pay $250m (RM1.07bn) over the next ten years to acquire the company's chip design plans.
This initiative is part of the country’s plan to produce its own graphics processing unit chips within five to ten years, capitalising on the growing demand for AI and data centre technologies.
The government aims to establish Malaysia as a hub for AI chip production, with plans to design, manufacture, test, and assemble these chips domestically and market them globally.
The partnership with Arm Holdings is structured around three core components. Firstly, it will launch training programmes for 10,000 integrated circuit design engineers, aiming to strengthen the country's semiconductor talent pool.
Selected Malaysian companies will gain privileged access to Arm's technology and intellectual property portfolio.
The collaboration will also focus on supporting the development of locally designed semiconductor products.
In addition to providing advanced semiconductor design technologies, Arm Holdings will also set up its first ASEAN office in Kuala Lumpur.
Malaysian Economy Minister Rafizi Ramli was cited by Reuters as saying that Malaysia will pay for access to seven high-end chip design blueprints from Arm.
Ramli highlighted the government's aspirations for this deal to enable local producers to scale up, with the objective of creating ten local chip companies each generating annual revenues between $1.5bn and $2bn.
In January 2025, Reuters reported that SoftBank Group and its majority-owned Arm Holdings were in discussions to acquire Ampere Computing, a semiconductor company backed by Oracle.
Ampere Computing, which utilizes Arm technology, produces central processing chips that are employed by companies such as Oracle and Google's Alphabet.
These chips are designed to be more energy-efficient compared to those from Intel and Advanced Micro Devices.