Intel is making headlines again—not for a new chip, but for potentially divesting a major portion of its networking business. Sources close to the matter suggest that Intel is actively exploring a sale of its network unit, signaling a deeper commitment to its strategic refocus on foundry services and artificial intelligence (AI).

This move, while not entirely unexpected, could mark a turning point in the global telecom infrastructure and semiconductor landscape. With multiple high-profile players reportedly in the running—including Broadcom, Nokia, Ericsson, private equity firms, and possibly even Chinese vendors like Huawei or ZTE—the implications are far-reaching.

Let’s break down what’s happening, who might buy it, what the deal could be worth, and why this matters for the industry

🔍 What Is Intel Selling?

Intel’s networking business includes several assets that support telecommunications infrastructure, cloud networking, and edge computing solutions. This spans software-defined networking (SDN), 5G radio access network (RAN) components, Ethernet controllers, and infrastructure processing units (IPUs). A notable component is Intel’s FlexRAN platform, a software-based 5G solution used by operators to virtualize RAN components.

Estimated Annual Revenue:

Intel’s network and edge group (NEX), which encompasses these assets, generated $7.2 billion in revenue in 2023, down from $8.9 billion in 2022, reflecting market headwinds and strategic de-emphasis.

While not directly confirmed, analysts estimate that the portion Intel may divest represents $3.5 to $4.5 billion in annual revenue, with operating margins in the low single digits—significantly below Intel’s corporate average.

🏁 Who Are the Potential Buyers?

1. Broadcom

Broadcom is perhaps the most aggressive acquirer in the semiconductor space. With over $60 billion in acquisitions in the last decade—including CA Technologies, Symantec Enterprise, and most recently VMware (acquired for $69 billion in 2023)—Broadcom has the capital, appetite, and integration experience to absorb Intel’s network assets.

Why it fits:

Broadcom’s push into AI and edge connectivity could benefit from Intel’s telco footprint and IP portfolio, especially in programmable network solutions.

Estimated Bid Value: $6–7 billion

2. Nokia

Nokia, once a mobile phone titan, has reshaped itself into a global telecom infrastructure giant. Its strength in 5G, optical networks, and private wireless makes Intel’s assets a natural complement.

Why it fits:

Nokia could integrate Intel’s FlexRAN and IPU stack into its 5G base station and Open RAN efforts, enhancing cost-efficiency and market competitiveness.

Estimated Bid Value: $5–6 billion

3. Ericsson

Ericsson, a peer of Nokia, is similarly invested in next-gen telecom gear. With its focus on cloud-native infrastructure, virtual RAN (vRAN), and Open RAN, acquiring Intel’s software-defined assets would deepen its capabilities.

Why it fits:

The synergy between Intel’s programmable network silicon and Ericsson’s software stack could create an end-to-end 5G edge solution.

Estimated Bid Value: $4.5–5.5 billion

4. Private Equity Firms

If strategic buyers falter—due to regulatory or financial hurdles—private equity could step in. Firms like Silver Lake, KKR, or Bain Capital have previously invested in infrastructure-heavy and semiconductor-adjacent businesses.

Why it fits:

PE firms could carve out Intel’s assets, restructure for efficiency, and re-sell to strategic buyers in 3–5 years. The challenge would be managing talent and IP-heavy operations without strong domain expertise.

Estimated Bid Value: $4–5 billion

5. Chinese Vendors (Huawei, ZTE)

Despite geopolitical barriers, it’s worth noting that Huawei and ZTE might view Intel’s network assets as a way to bolster domestic self-reliance. However, U.S. export controls and national security concerns make this outcome highly unlikely.

Why it fits (in theory):

Huawei is investing aggressively in RISC-V, chip design, and AI accelerators. Intel’s network IP could theoretically accelerate this.

Reality Check:

Regulatory approval in the U.S. is next to impossible. Any involvement would draw immediate national security objections.

Estimated Bid Value (hypothetical): $6–8 billion, but politically nonviable.

🔄 Why Is Intel Selling

Intel’s CEO Pat Gelsinger has made it clear: the company is doubling down on its foundry ambitions (IFS) and AI leadership. Shedding non-core or underperforming assets is a part of this restructuring journey.

In the Q2 2025 earnings call, Intel noted that capital allocation will increasingly favor:

1. High-margin AI compute (like Gaudi and Falcon Shores platforms)
2. Expansion of Intel Foundry Services (IFS), with clients like Microsoft and OpenAI
3. Cutting operational drag from businesses with low ROI or margin pressure

Capital Reallocation Impact:

If Intel nets $5–6 billion from this sale, it could redirect those funds toward building out 2nm and sub-2nm foundry capacity or acquiring smaller AI startups to strengthen its software stack.

Industry Implications:

1. Telecom Consolidation

If Ericsson or Nokia acquires the assets, it will further consolidate the telecom equipment space—potentially impacting vendor pricing, innovation pace, and global supply dynamics.

2. AI-Networking Convergence

Broadcom or even private equity could reposition the business to serve the emerging AI-networking edge, especially as data centers shift toward disaggregated architectures.

3. Global Tech Geopolitics

Any Chinese bid will likely face strong opposition from U.S. and EU regulators. But even without China, the deal will draw scrutiny from the Committee on Foreign Investment in the U.S. (CFIUS), especially if sensitive IP or telco infrastructure assets are involved.

4. Talent and IP Transfer

This is not just about hardware. Intel’s networking group includes hundreds of software and hardware engineers with deep knowledge in programmable silicon, vRAN, and cloud-native stacks. The acquirer gains access to strategic IP and technical talent that could influence standard-setting across the telecom industry.

📉 Risks and Unknowns

1. Regulatory Overhang: Any deal over $2B involving critical infrastructure will be closely reviewed by U.S. and international regulators.
2. Integration Complexity: Intel’s network assets are deeply integrated into its broader silicon and software stack. Unwinding or separating them for a clean sale won’t be trivial.
3. Cultural Mismatch: A financial buyer (e.g., PE) may struggle to retain engineering talent without a compelling long-term product roadmap.

💡 Final Thoughts

Intel’s network business sale is more than a corporate reshuffling. It’s a bellwether for where the industry is heading — toward tighter alignment between AI compute, edge infrastructure, and disaggregated telecom. Whoever ends up acquiring this unit isn’t just buying revenue—they’re buying a seat at the table in the next wave of networked intelligence.

As AI workloads push toward the edge and telcos hunt for more flexible, software-defined architectures, this deal could shape the architecture—and ownership—of the next-generation internet.

📚 Sources

1. Intel Q2 2025 Earnings Call
2. Intel Annual Report 2023
3. Bloomberg, “Intel Considers Sale of Networking Unit Amid AI Shift” (June 2025)
4. Reuters, “Broadcom and Nokia Named Among Potential Buyers of Intel’s Network Division” (July 2025)
5. IDC Worldwide Telecommunications CapEx Forecast 2025–2028
6. Company Financial Disclosures (Broadcom, Nokia, Ericsson 10-Ks)