Saturday, October 24, 2026
AI Infrastructure · News & Analysis
Commentary

Cambricon Q3 2026 Results: Profit Doubles as Trillion-Yuan Valuation Fades

Cambricon's cumulative profit for the first nine months of 2026 reached 3.7 billion yuan with third-quarter earnings doubling year-on-year, but the trillion-yuan market capitalization that once marked the stock's peak has not been sustained.

Cambricon's third-quarter results, reported in October 2026, mark a notable earnings inflection: profit for the period doubled year-on-year, bringing cumulative earnings for the first nine months of 2026 to 3.7 billion yuan. For a company that spent years as a pre-profit pioneer in dedicated AI chip design before listing on Shanghai's STAR Market in 2020, that figure represents genuine operational progress. That it arrives alongside a significantly cooler market valuation — the trillion-yuan market capitalization that once defined Cambricon's peak is, by all accounts, now firmly in the past — captures the central tension facing China's leading homegrown AI accelerator.

Cambricon shares closed at 932.2 on October 22, up 5.5% on the session; no report reviewed by Slicast specifically identifies the catalyst for that move. The broader valuation trajectory, however, signals something structural. At its peak, the trillion-yuan market cap embedded an assumption that domestic AI chip design could command multiples comparable to global semiconductor franchises. That premium has since compressed — a pattern familiar to investors who have tracked similar re-ratings across AI infrastructure verticals once the initial enthusiasm cycle matures and competitive realities set in.

The fundamental investment case for Cambricon remains substantive. Analysts cited in August 2026 reports projected that China's homegrown AI accelerator makers would collectively supply approximately 90% of the domestic market, with Cambricon and Huawei identified as the expected primary beneficiaries of the accelerating structural shift away from Nvidia and AMD. That projection, if borne out, would represent a near-complete reshaping of China's AI infrastructure supply chain over a compressed timeline. The policy environment has been a persistent enabler: as reported in August 2025, Beijing's initiative to mandate homegrown AI chips for domestic datacenters provided both a demand floor and a fundraising environment in which Cambricon reportedly secured approximately $560 million to sharpen its competitive position.

The competitive landscape, however, is filling in rapidly. As reported in September 2026, four domestic GPU makers — 燧原科技, 沐曦, 摩尔线程, and 壁仞科技 — collectively entered capital markets at a combined valuation exceeding 613.2 billion yuan, signaling the opening of a more crowded phase of China's domestic AI chip buildout. Cambricon's market leadership, cited alongside Huawei in June 2026 reporting on the pair actively capturing Nvidia's retreating China market share, was built when the domestic field was comparatively sparse. Sustaining that position as well-capitalized GPU specialists scale will require continued architectural differentiation and ecosystem depth that cannot be assumed.

The overall picture is of a company navigating the transition from an early-stage, high-multiple story toward a more mature earnings-driven investment case. The 3.7 billion yuan nine-month profit figure, with third-quarter profit doubling year-on-year, suggests the operating model is maturing at pace. Three concrete signals are worth tracking: whether the domestic procurement mandate can sustain demand momentum into 2027 or whether lumpy government procurement cycles introduce revenue variability; how Huawei's own chip ambitions evolve relative to Cambricon, given that the two are currently co-beneficiaries of the same structural trade; and whether the four newly listed GPU challengers can translate capital-market momentum into credible architectural competition within the datacenter workload segment. The distance between Cambricon's current trading price and its trillion-yuan peak is a reminder that in AI infrastructure, market valuations price expectations — and expectations have a way of running well ahead of even genuinely impressive results.

Based on 6 archived reports · Cambricon →
Cambricon Q3 2026 Results: Profit Doubles as Trillion-Yuan Valuation Fades · Slicast