In November of this year, it was reported that Intel was interested in acquiring the company for $5 billion.AI startup SambaNovaFollowing reports that RDU architecture design would accelerate AI transformation, another development has emerged.The actual transaction amount may shrink to $16 billion..
According to a Bloomberg News report, compared to previous reports of a $50 billion acquisition of SambaNova, this report suggests that the deal may be worth only around $16 billion.
If this price is accurate, it would undoubtedly be a "bloodbath" for SambaNova. This startup, founded in 2017, was valued at over $50 billion when it raised funds in 2021. This reflects the fact that second-tier AI chip startups are facing survival pressure and valuation corrections in the current AI hardware market where the strong get stronger (NVIDIA dominates).
CEO Chen Liwu is the key driving force
The role of Intel's current CEO, Lip-Bu Tan, behind this potential deal is intriguing.
SambaNova already boasted a prestigious investor lineup, including SoftBank, Google Ventures, and Intel's own Intel Capital. Chen Liwu's personal investment firm, Walden International, was also an early investor, and Chen himself was appointed as an executive director of SambaNova last May. This close relationship is seen by outsiders as a key catalyst that accelerated negotiations between the two parties.
What to do with it? Targeting the "inference" and "complete system" markets.
SambaNova's core competitiveness lies in its DataScale system and SN40L processor, focusing on high-performance generative AI inference and fine-tuning. Unlike NVIDIA, which focuses on general-purpose GPU computing, SambaNova adopts a software-defined hardware architecture. Therefore, for Intel, acquiring SambaNova has several strategic implications:
• Enhance inference computing power:As the Gaudi series struggles to compete with the H100/B200 in the market, SambaNova technology can provide Intel with a differentiating advantage in specific areas, such as large language model inference.
• Complete solution:SambaNova doesn't just sell chips; it excels at providing complete solutions that include hardware, compilers, and model-as-a-Service, which is exactly the piece of the puzzle that Intel is currently missing in its attempt to transform into a "system foundry" and "enterprise AI service".
Analysis: Mergers and acquisitions are easy, but integration is difficult.
In my opinion, Intel's move is clearly aimed at quickly closing the technological gap with its competitors. $16 billion is not a small sum for Intel, which is currently facing financial difficulties, but it may be a gamble it has to take.
However, Intel's past acquisition record in the AI field has not been perfect (for example, the painful integration period of Nervana and Habana Labs). If they can successfully acquire SambaNova this time, the biggest challenge for Chen Liwu's team will be how to smoothly integrate its unique architecture into Intel's existing OneAPI ecosystem without crowding out its own Gaudi product line.
Intel and SambaNova have declined to comment on the news. The deal could be finalized within weeks, but could be delayed due to due diligence.
