Intel's SambaNova acquisition was seen as a way to spearhead the company's AI strategy in inference, but it appears Team Blue has settled for much less.
Intel's CEO Now Intends to Invest Directly in SambaNova's Funding Round After Previously Backing It Through Walden Capital
When we talk about Intel, the company is one of the only compute providers that hasn't managed to capitalize on the AI frenzy at all, and this has been a persistent problem for several quarters. Intel's former CEO, Pat Gelsinger, has acknowledged the lack of intent in AI, and it appears that missing out on the SambaNova opportunity is another example. Based on Intel's recent blog post, the firm is entering into an agreement with SambaNova for "Xeon-based" infrastructure, along with offering an inference solution that combines Intel and SambaNova components.
This collaboration complements Intel’s existing data center GPU commitments and does not alter its path forward to competing in AI. The company continues to invest across GPU IP, architecture, products, software, systems and strengthen its roadmap as part of its edge-to-cloud AI engagements.
Together, Intel and SambaNova aim to help shape the next generation of heterogeneous AI data centers, integrating Intel Xeon processors, Intel GPUs, Intel networking and storage, and SambaNova systems—to unlock a multi-billion-dollar inference market opportunity.
The idea behind SambaNova is simple. The company has specialized processing engines called the Reconfigurable Data Unit (RDU), which map entire neural network graphs directly into hardware, similar to what Etched or Taalas AI is doing. SambaNova recently unveiled its SN50 AI chip, revealing that it has achieved 3x lower costs than GPUs in agentic AI workloads, along with 5x more compute per accelerator than the previous generation. For Intel, the plan was to acquire SambaNova and speed up the time to market for an inference-focused option, but it seems the direction is different.
SambaNova is also raising $350M in investment in the Series E round, backed by the likes of SoftBank and Intel Capital, and, interestingly, Lip-Bu Tan is already an early SambaNova investor through his Walden Capital portfolio. So ultimately, Tan has two positions open up at SambaNova, which means that the CEO is confident in the company's ability to capitalize on the inference hype. But, now that an acquisition is ruled out, what could be next?
Well, our estimates suggest that Intel might integrate SambaNova's RDUs into specific workloads, similar to what NVIDIA intends to do with Groq. Team Blue has already hinted at this in its latest blog post, and the company plans to build "heterogeneous" AI data centers with Intel GPUs, CPUs, and SambaNova systems. But when we could see such a solution come to fruition, it would be a lot more critical, considering that Intel has already delayed the training hype, and losing in inference would be a costly mistake.
About the author: Muhammad Zuhair is a hardware and technology reporter for Wccftech, specializing in the semiconductor industry and the complex interplay between technology, manufacturing, and geopolitics. His coverage focuses on the corporate strategies and technological roadmaps of industry giants like TSMC, NVIDIA, Samsung, and Intel. Zuhair's expertise lies in deconstructing complex topics such as fabrication nodes (e.g., 2nm process), the economic impact of policies like the CHIPS Act, and the strategic development of AI infrastructure from NVIDIA, AMD and Intel.
Follow Wccftech on Google to get more of our news coverage in your feeds.
