When Qualcomm announced in June 2025 that it would buy Alphawave for US$2.4 billion, the headline was a milestone for a company most people had never heard of. Inside the semiconductor world, however, the name had been known for years, and the man behind it was already on his second company.
Tony Pialis had earlier founded V Semiconductor, a chip-design firm, which Intel went on to acquire. At Intel he served as a vice president running analog and mixed-signal intellectual property, work that sits deep inside the chips that move data at high speed. He left that role to start again, and in 2017 he co-founded Alphawave in Toronto.
Customers can take years to design a product, and revenue lands only after they commit.
The idea was precise. Data centres were growing faster than the wires that connect them. Every chip, from a graphics processor to a networking switch, needs a way to send data across a board or a rack at enormous speeds. Designing those links is expert work, and few chip companies want to hire a team just to do it. Alphawave sold them the finished design to license, along with later chips and chiplets built around it.
That licensing model carries a particular kind of risk. Customers can take years to design a product, and revenue lands only after they commit. A company in that position needs patience and a technical reputation strong enough to win design slots from large firms. Alphawave built that reputation quickly enough to list on the London Stock Exchange in 2021, an unusual choice for a Toronto-based technology company, since most Canadian founders aim for Toronto or New York.
The real test came with the rise of artificial intelligence. Training large models depends on moving huge quantities of data between processors, and demand for exactly the kind of connectivity Alphawave made surged. That demand caught the attention of Qualcomm, which was looking to build out its data-centre business. The agreed price was US$2.48 per share in cash, or about 183 pence, and the board backed it unanimously.
Qualcomm completed the purchase in December 2025, about a quarter ahead of schedule. Pialis stayed on, and was named to lead Qualcomm's data-centre business, a rare outcome in which a founder of an acquired company is handed a major operating role rather than a farewell package.
The pattern in his career is worth noticing. One company sold to Intel, a second to Qualcomm, each in a niche most investors find intimidating. The lesson for an Indian founder building deep technology is that you do not need a consumer brand. You need a hard problem, a customer base that cannot easily replace you, and the stamina to solve it more than once.