AMD to acquire Taalas in second Canadian AI chip deal
What's the deal? AMDDealroom has a profile for this one. Try Dealroom → has reached a definitive agreement to acquire Toronto-based Taalas, a startup that hardwires AI models directly onto the chips that power them. Financial terms were not disclosed, and the deal remains subject to closing conditions and regulatory approval.
What does Taalas do? Founded in 2023, the startup uses hard-wired computation to convert AI models into custom silicon that can replace general-purpose graphics processing units (GPUs). It emerged from stealth in 2024 with $50 million in funding from Quiet Capital and Pierre Lamond, among others.
Who's behind it? Taalas was started by a trio of former AMD employees and Tenstorrent leaders, including chief executive officer Ljubisa Bajic, a Tenstorrent co-founder and former CEO. Its team also includes chief operating officer Lejla Bajic and chief technology officer Drago Ignjatovic.
The strategic rationale: AMD called Taalas "a pioneer in specialised AI inference silicon." Its technology "strengthen[s] our AI portfolio by delivering differentiated inference performance and efficiency," said Vamsi Boppana, senior vice-president of AMD's AI group.
What's the endgame? Taalas claims it can launch new chips in about two months, faster than the industry standard, producing models it says are a thousand times more efficient than software counterparts. It is betting single chips could outperform small GPU data centres.
Why AMD? "Joining AMD will give us the scale, engineering resources and global reach to accelerate our innovation," Bajic said. AMD has operated in Canada since its 2006 purchase of Markham-based graphics chipmaker ATI TechnologiesDealroom has a profile for this one. Try Dealroom →.
The signal: This is AMD's second acquisition of a Canadian AI chip firm in just over a year, following its 2025 deal for the team behind Toronto's Untether AI. As demand for AI compute grows, chipmakers are racing to lock up specialised inference talent and technology that could challenge GPU dominance.
Read more: BetaKit
Image credit: geerlingguy
