Intel vs TSMC Q2 2026: Two Very Different Semiconductor Stories

The semiconductor world's two most-watched companies reported Q2 2026 earnings weeks apart, and the results tell a story of dominance versus recovery. TSMC delivered a record quarter with $40.2 billion in revenue, 67.7% gross margins, and a net profit that alone exceeded Intel's entire revenue. Intel, meanwhile, posted $16.1 billion in revenue — its fastest growth in 15 years — yet reported an $11 billion GAAP net loss that left casual observers scratching their heads. The truth, as always, is more nuanced than the headline numbers suggest.

Q2 2026 Head-to-Head Financial Snapshot

Metric	TSMC	Intel
Revenue	$40.2B (+33.7% YoY)	$16.1B (+25% YoY)
Gross Margin	67.7% (above 65.5-67.5% guidance)	41.8% non-GAAP
Operating Margin	60.3%	Foundry alone: -$2.1B operating loss
Net Income (GAAP)	~$22B net profit (+77% YoY)	-$11.0B GAAP net loss
Non-GAAP Net Income	N/A	+$2.2B ($0.42 EPS)

The scale gap is stark. TSMC's Q2 revenue ($40.2B) is 2.5x Intel's ($16.1B), and its net profit alone (~$22B) exceeds Intel's total revenue. Advanced nodes (7nm and below) contributed 77% of TSMC's wafer revenue , while Intel's product revenue growth is being driven by a broad recovery, not yet by foundry leadership.

The $11 Billion Accounting Artifact: What Really Happened at Intel

Intel's GAAP net loss of $11.0 billion sounds catastrophic, but it is almost entirely an accounting artifact. The loss was driven by a $12.5 billion non-cash mark-to-market charge on escrowed shares held for the CHIPS Act agreement with the U.S. Department of Commerce . As Intel's own share price rose, accounting rules forced Intel to record the increased value of those escrowed shares as a loss — a counterintuitive dynamic that has nothing to do with its operating performance .

On a non-GAAP basis, Intel generated $2.2 billion in net income, marking its seventh straight quarter of beating its own financial guidance . CEO Lip-Bu Tan told investors the quarter represented "the strongest revenue growth in more than 15 years" .

TSMC: Record Margins and First 2nm Revenue

TSMC is operating from a position of overwhelming financial and technical strength. Gross margin expanded 150 basis points sequentially to 67.7%, surpassing the high end of its guidance range and marking a 910-basis-point improvement from the prior year . Operating margin reached 60.3%, and return on equity hit an all-time high of 45.9% .

High-performance computing (HPC) remained TSMC's growth engine, surging 20% quarter-over-quarter to account for 66% of Q2 revenue . Smartphone revenue declined 4% to 22% of revenue, while IoT contributed 8% .

The headline strategic milestone: TSMC's 2nm process contributed 3% of wafer revenue in Q2 2026 — its first disclosed commercial quarter . Management expects the 2nm ramp to subtract approximately 3-4 percentage points from gross margin in the second half of 2026 as initial production costs are absorbed .

Intel's Recovery: Product Strength, Foundry Weakness

Intel's Q2 2026 revenue of $16.1 billion beat analyst consensus of roughly $14.43 billion by about 12%, and non-GAAP EPS of $0.42 nearly doubled the $0.21 estimate . All three reporting segments grew: Client Computing & Physical AI was up 13% YoY, Data Center and AI revenue surged 59% YoY, and Foundry revenue grew 31% to $5.8 billion .

But the Foundry segment remains a deep financial drag. It posted an operating loss of $2.1 billion, though this narrowed from -$2.4 billion in Q1 2026, as higher yields, improved cycle times, and increased factory scale across Intel 4, Intel 3, and 18A drove better wafer costs .

The Process Node Race: TSMC's 2nm vs Intel's 18A

TSMC is already shipping 2nm commercially; the process contributed real — if early — revenue in Q2 2026. Intel's 18A node entered production in 2025, with its first client chip (Panther Lake, Core Ultra Series 3) launching in January 2026 . The enhanced 18A-P entered risk production as of June 2026, claiming 9% higher performance at iso-power or 18% lower power at iso-performance compared to baseline 18A .

Intel's most significant external validation came in March 2026, when Nvidia selected Intel's Xeon 6 processor as the host CPU for its DGX Rubin NVL8 systems, announced at GTC 2026 . Each DGX Rubin NVL8 system features two Xeon 6776P CPUs alongside eight Rubin GPUs for large-scale AI workloads . However, this is a CPU slot in Nvidia's AI server — a product win for Intel's Data Center and AI group — not a foundry win for the 18A process.

The critical question for Intel's foundry ambitions remains unanswered: Can 18A win external chip design wins from Nvidia or other major fabless firms? Reports that Intel is working to finalize Nvidia as a foundry customer for 18A remain unconfirmed and speculative .

Financial Vulnerability and Strategic Risk

While Intel's core product business is recovering, financial vulnerability remains high. The Foundry segment burns $2.1 billion per quarter, and the CHIPS Act escrow accounting creates persistent GAAP volatility that distorts reported results . Intel's $20+ billion annual capex program relies on a combination of CHIPS Act grants, potential external investment, and operating cash flow.

TSMC, by contrast, is self-funding its massive expansion from record operating profits. Its guidance for full-year 2026 raised expectations to approximately 40% growth as AI demand continues to overwhelm supply .

Bottom Line

TSMC is the undisputed champion of the Q2 2026 period: record revenue, industry-leading margins, and a technology roadmap (2nm already commercial) that keeps it years ahead. Intel showed real operational momentum under CEO Lip-Bu Tan — 25% revenue growth, a seventh straight beat, and the Xeon 6 DGX Rubin win — but remains a turnaround story. Its foundry business burns $2.1 billion per quarter, its $11 billion GAAP net loss from accounting mechanics confuses the market, and the single most important question — whether 18A can win external foundry customers — is still ahead of it.