Intel Said to Debate SiFive Buyout to Bolster Chip Technology Against Arm

Intel is debating a possible offer to buy SiFive, a person familiar with the matter told Reuters, a company closely associated with open-source technology that is challenging the rise of Intel's rival, Arm.

SiFive, a San Mateo, California-based startup, employs several of the creators of RISC-V, an open-source chip technology that is challenging Arm, the British chip technology firm being acquired by Nvidia for $40 billion (roughly Rs. 10,21,640 crores). Both Arm and SiFive sell intellectual property such as chip designs to others who ultimately produce the chips.

Intel and SiFive both declined to comment.

Bloomberg on Thursday reported Intel's interest, citing a source saying that the Santa Clara, California-based chipmaker is mulling a $2 billion (roughly Rs. 14,590 crores) offer. Intel, along with rivals such as Qualcomm, is already an investor in SiFive, which raised $61 million (roughly Rs. 450 crores) in a funding round led by Korea's SK Hynix.

SiFive is designing computing cores using the RISC-V architecture. While the underlying architecture for those cores is open-source, the specific core designs themselves can be sold.

Purchasing SiFive could give Intel a library of intellectual property it could use both in its own chips and that it could offer to license to future customers as it works to build a business by opening up its chip factories to outsiders. Intel has already said it plans to license out computing cores based on its own proprietary x86 architecture to customers as part of its contract manufacturing business.

But Intel would also gain a software boost. SiFive is also working on making it easier to programme to different kinds of computing chips and last year hired Chris Lattner, a prominent Silicon Valley computer scientist.

Lattner spearheaded the creation of the Swift programming language for Apple that has become the primary way developers write apps for iPhone models. More recently, Lattner oversaw programming language teams for Alphabet's Google Brain and TensorFlow artificial intelligence teams.

© Thomson Reuters 2021