Huang calls $5B Intel stake ‘incredible investment’

Nvidia CEO Jensen Huang revealed that the company’s substantial $5 billion investment and extensive technology collaboration with Intel followed nearly a year of in-depth discussions between the two tech giants.

Huang personally engaged with Intel CEO Lip-Bu Tan to forge this strategic alliance. Describing Tan as a “longtime friend,” Huang addressed reporters on Thursday following the unveiling of Nvidia’s co-development initiative with Intel, centered on creating cutting-edge data center and PC chips. This collaboration is a cornerstone of the broader investment agreement. Tan echoed this sentiment, stating their professional relationship spans three decades.

“We envisioned this as a truly game-changing investment,” Huang stated, emphasizing the potential for synergistic innovation.

Nvidia is set to partner with Intel in developing advanced artificial intelligence (AI) systems tailored for data centers. These systems will integrate Intel’s x86-based central processing units (CPUs) with Nvidia’s high-performance graphics processing units (GPUs) and advanced networking solutions, aiming to deliver unparalleled computational power for AI workloads.

The collaboration also extends to the PC market, where Intel will incorporate Nvidia GPUs into its CPUs for PCs and notebooks. This move is anticipated to deliver enhanced graphics capabilities and cater to the growing demand for AI-powered applications on personal computing devices.

Intel’s revenue chief, Greg Ernst, noted in a recent post that the intricacies of the deal took several months to finalize, culminating in an agreement reached just days prior to the official announcement. The swift finalization speaks to the strategic imperative and mutual benefit both companies see in the collaboration.

This substantial investment underscored a significant power shift within Silicon Valley, fueled by the surging demand for AI solutions that gained momentum after OpenAI’s launch of ChatGPT in late 2022. The AI boom has dramatically reshaped the competitive landscape, with Nvidia emerging as a dominant force.

Nvidia’s valuation currently stands at over $4.25 trillion, a staggering contrast to Intel’s $143 billion. This divergence in market capitalization reflects Nvidia’s strategic positioning and technological prowess in the rapidly evolving AI sector. While Intel’s stock has seen a decrease of 31.78% over the last five years, Nvidia shares have skyrocketed by an impressive 1,348% demonstrating the divergent trajectories of the two companies amidst the AI revolution.

Traditionally, the CPU was the centerpiece of PCs and servers, with Intel commanding a dominant market share. However, the advent of AI infrastructure has shifted the paradigm. Modern AI systems, exemplified by the $4 billion data center Microsoft recently unveiled, now often require multiple Nvidia GPUs for every CPU, underscoring the growing importance of GPUs in AI processing.

Nvidia’s current high-end AI systems, such as the NVL72, are frequently powered by Arm-based CPUs rather than Intel’s x86 architecture. Huang emphasized that Nvidia plans to augment Intel’s CPUs within its NVLink racks for AI, which is a critical move that will foster a broader ecosystem and enable greater flexibility for customers.

“We will be procuring CPUs from Intel and integrating them into our superchips, essentially creating compute nodes that are integrated into rack-scale AI supercomputers,” Huang explained, highlighting the seamless integration of Intel’s CPUs into Nvidia’s complex compute architecture.

Nvidia will also be providing GPU technology for Intel chips destined for laptops and PCs. This is a calculated move into a market segment that Huang believes has been underserved. He estimates that the addressable markets for these collaborations are worth an impressive $50 billion in total, a figure that underscores the immense potential for both companies.

“We are poised to become a very significant customer of Intel’s CPUs, and concurrently, we will be a substantial supplier of GPU chiplets for incorporation into Intel’s chips,” Huang elaborated, signaling a symbiotic relationship with significant mutual benefits.

Huang reassured that the partnership with Intel would have “no” adverse impact on Nvidia’s ongoing relationship with Arm, suggesting that the collaboration is complementary rather than competitive with their existing partnerships.

The investment deal primarily focuses on the integration of Nvidia and Intel products and does not extend to foundry services for now, However, the two companies did not discount future partnerships in this field, suggesting that the current deal is the foundation for further strategic alignment.

“We consistently evaluate Intel’s foundry technology and plan to continue doing so. However, this announcement is specifically centered on our custom CPUs,” Huang clarified. Currently Nvidia uses Taiwan Semiconductor Manufacturing Company (TSMC) to manufacture the majority of its chips, implying that potential future foundry services might offer redundancy and flexibility in the supply chain.

The collaborative efforts will leverage Intel’s advanced packaging technology, which is a phase of chip manufacturing that combines multiple chip components into a single, integrated unit. This contributes to the performance and efficiency of the chip and allows for greater product integration.

Tan expressed his appreciation for Nvidia’s support and confidence in Intel’s capabilities.

“I extend my gratitude to Jensen for his confidence in both myself and our team, and Intel will be fully committed to delivering substantial returns on this investment,” Tan stated, reiterating the seriousness of the commitment to Nvidia.

The preceding year had seen Intel’s board make the decision to replace former CEO Pat Gelsinger amidst escalating costs in its manufacturing operations and a perceived failure to establish a presence in the burgeoning AI chip market. Last March, Intel appointed Tan, a renowned and exceptionally well-connected investor with a successful track record of revitalizing chip software firm Cadence Design Systems, as its new CEO.

Since assuming the role, Tan has prioritized cost reduction and fundraising initiatives at Intel, even as the future of Intel Foundry, the company’s chip manufacturing arm, remains under scrutiny. These initiatives are strategic moves to restore financial stability and market confidence.

Besides the $5 billion financial infusion from Nvidia and the $8.9 billion in government assistance, Intel has also secured a $2 billion investment from SoftBank, divested a majority stake in its ASIC subsidiary Altera to Silver Lake for $3.3 billion, and sold $1 billion worth of stock from Mobileye, its autonomous vehicle subsidiary.

Intel has also undertaken significant workforce reductions, announcing in July that it intends to eliminate 15% of its staff by the end of the year. These cuts are part of the company’s broader restructuring plan to improve resource allocation and streamline operations.

Intel both designs and manufactures its own chips, but it has ambitions to expand its business to include manufacturing chips for other companies in a similar fashion to pure-play foundries such as TSMC or Samsung. While Intel is looking to secure major players such as Nvidia or Apple as customers, analysts suggest that a major client is vital to demonstrate that their technology is mature and ready for mass production. Without such a client, Intel may struggle to convince other firms that its manufacturing is up to par.

However, state-of-the-art chip manufacturing comes at a steep price, and Intel has indicated that if it is unable to secure enough customers, it may curtail its investments in its foundry businesses. A scenario of this kind may prompt reactions from government officials; Intel is seen to be of paramount importance for the nation’s strategic interests given that it is the sole American entity that is capable of manufacturing the cutting-edge chips needed for both defense and communication applications.

The Trump administration secured a 10% stake in Intel in August. Intel was previously slated to receive $8.9 billion in grants and loans from the CHIPS Act, however, the Trump administration requested, and received, an equity stake in the chipmaker in return for those funds. This strategic move highlights the government’s focus on maintaining domestic control over chip production and protecting domestic interests.

Huang was attending a State Dinner at Windsor Palace with Trump this week and he and Trump were announcing projects and investments in the UK. As such, a White House official and Huang assured CNBC that the Trump administration was not involved in this recent deal.

“Intel’s new partnership with Nvidia is a significant milestone for American high-tech manufacturing,” Kush Desai White House spokesman said in a statement.

— CNBC contributed to this story