Hopes this will improve its margin and profitability
Tech giant posts fastest growth in cloud and AI revenue in 22 quarters, with capex up 75% amid fierce competition in the sector.
Alibaba deploys Zhenwu M890 (3× performance) to replace Nvidia, using 470k T-Head shipments to lift cloud margins from 34.5%. Vertical integration amid US GPU bans forces competitors toward in-house semiconductors, reshaping AI infrastructure competition and pricing.
Alibaba Cloud cuts AI server payback to 2.5 years using proprietary chips versus Nvidia, achieving 45% cloud growth. This vertical integration marks Alibaba's pivot from e-commerce to AI as primary growth engine with improved margin economics for competitive pricing.
T-Head Gen2 di Alibaba raggiunge tape-out H2 2026 per large-model training; Zhenwu M890 già in scaled sales 650+ clienti. Stack silicon verticale cinese (GPU/CPU/networking) minaccia vendor AI globali; CTO ricalcolano TCO infra cloud e supply-chain risk.
Alibaba's AI Cloud division grew 45% YoY; AI product revenue hit 12.4B yuan, but capital expenditure surged 75% to 67.7B yuan. Hyperscalers are sacrificing profitability (EBITDA down 30% YoY) to lock in AI infrastructure scale—a capital-intensive race smaller players cannot afford to join.
Alibaba's AI cloud surged 44.9% YoY but capex doubled to $10B, EBITA fell 29.6%, free cash flow -$6.6B. IT leaders note: scaling AI infrastructure eats 25% of quarterly revenue while Qwen losses widen—profitability lags despite AI product growth.
Alibaba expects improved gross margins and profitability by deploying T-Head chips in AI data centers, with 470,000 units shipped and cloud
Cloudy AI is Chinese giant’s ‘most certain’ path to growth as e-commerce slows
Alibaba CEO Eddie Wu said on Aug. 20 that the company’s second-generation T-Head chip is expected to begin tape-out and production in the second half of
Faster cloud growth is being accompanied by mounting pressure on earnings and cash flow from AI infrastructure spending.