AI Infrastructure Splits the Chip Market: Teradyne, Cadence, and ASM Raise Guidance While Consumer Demand Falters

Weakness revealed:

• Teradyne's memory test TAM declined 4% year-over-year in 2025, with NAND at "low levels" — only AI-driven HBM and DRAM segments grew .
• The company flagged "continued order pushouts from auto and industrial customers" due to macroeconomic uncertainty, and Robotics posted a material operating loss .

Cadence Design Systems: Three Guidance Raises in Six Months

How guidance was raised: Cadence raised its full-year 2026 revenue outlook three times in six months. Starting from an initial FY2026 guide of $5.9 B–$6.0 B in February 2026 , it raised to $6.125 B–$6.225 B in April , and again to $6.26 B–$6.34 B in July after a strong Q2 . Management cited "robust demand for AI-powered chip and system design software" from chipmakers and tech companies developing increasingly sophisticated AI processors .

Key guidance figures:

• FY2026 revenue guidance raised to $6.26 B–$6.34 B (18%–20% growth) .
• Q2 2026 revenue: $1.58 B ; Q1 2026: $1.474 B (+18.7% YoY) .
• IP business grew >40% YoY, driven by AI and HPC applications .

Weakness revealed: The available sources for Cadence do not contain direct commentary on consumer or automotive end-market weakness. As a pure-play EDA (electronic design automation) and IP licensing company, Cadence is largely insulated from end-market chip demand fluctuations — its customers are chip designers, not automotive or consumer OEMs. The weakness in those markets is thus implicit in the broader semiconductor landscape rather than explicitly called out in Cadence's earnings.

ASM International: AI and China Rebound Fuel Beats

How guidance was raised: ASM lifted its 2026 full-year outlook in March 2026, projecting Q1 2026 revenue of ~€830 M and stating Q2 would be higher, driven by AI investments and a China sales rebound . In April, it beat Q1 estimates with revenue of €863 M and issued a Q2 forecast ~10% above analyst expectations . On July 28, 2026, ASM forecast Q3 revenue of €1.1 B ($1.3 B), above analyst estimates, as "AI infrastructure build-out" continued to drive orders for its chipmaking equipment .

Key guidance figures:

• Q3 2026 revenue forecast: €1.1 B (~$1.3 B), above consensus .
• Q1 2026 revenue: €863 M (beat; Q2 guided higher) .

Weakness revealed: ASM's strong performance is explicitly tied to AI infrastructure and a China rebound. Non-AI end markets are not the driver. The company is also exposed to memory — its tools are used in deposition processes for DRAM and NAND — and the broader memory market outside AI-specific HBM remains soft. ASM itself noted that "AI-related spending has helped offset weakness in automotive, PC and memory markets" .

What the Divergence Reveals

The collective guidance from these three companies makes visible a structural divide in the semiconductor industry:

• AI infrastructure spending is overwhelming — Teradyne gets ~70% of revenue from AI, Cadence raised guidance 3x, and ASM consistently beats estimates. Data centers are projected to consume an estimated 70% of all memory chips produced by 2026 .
• Consumer, automotive, and industrial chip markets are weak — Teradyne explicitly cited order pushouts from auto and industrial customers, and its Robotics segment posted losses . The broader memory TAM (outside AI-driven HBM/DRAM) declined 4% in 2025 . ASM's China rebound aside, non-AI end-markets are not contributing meaningfully to the growth.
• The "AI vs. everything else" split is structural — a 2026 industry analysis describes a "semiconductor scarcity" where AI data centers have become the industry's priority customer, leaving automotive and consumer segments capacity-constrained .

Bottom line: The guidance figures show AI demand is strong enough to lift these companies to record revenue independently of the rest of the chip market. Consumer, automotive, and non-AI memory markets are not just lagging — they are actively contracting for some of these companies, creating one of the most pronounced bifurcations the semiconductor industry has seen.