The latest earnings season has revealed a stark contrast for China’s technology giants: while their revenue lines continue to climb, net profits are slipping under the weight of massive artificial intelligence (AI) investments. Bloomberg’s The China Show highlighted this trend in a recent video, noting that AI‑related capital expenditures have surged past 30 % of total R&D budgets for firms like Alibaba Group, Tencent Holdings, and Baidu Inc. This article dives into the numbers, examines the underlying drivers, and explores what the future may hold for China’s tech sector amid an AI‑fueled spending spree.

Link

Embed

Copy and paste this HTML code into your webpage to embed.

১. AI খরচের বোmbা: কতটা বাড়ছে?

According to a Counterpoint Research report released in Q1 2026, Chinese tech conglomerates collectively allocated US$84 billion to AI infrastructure in FY 2025, a 42 % year‑over‑year increase. The spending breakdown shows:

• Semiconductor procurement (AI accelerators, HBM memory): 38 %
• Data‑center expansion and cooling: 27 %
• AI talent acquisition and salaries: 20 %
• Software frameworks and model licensing: 15 %

These figures outpace the growth of traditional cloud services, which rose only 12 % in the same period. The aggressive push is driven by Beijing’s “New Generation AI Development Plan” (2021‑2030), which earmarks ¥1 trillion for AI breakthroughs by 2030, prompting firms to front‑load investments to secure policy subsidies and market leadership.

২. আয়ের Numbers: কী দেখানো হচ্ছে?

Despite revenue growth of 9‑11 % across the big three, net profit margins have contracted sharply:

Alibaba’s FY 2025 net profit fell to ¥84 billion from ¥112 billion the prior year, a 25 % drop, while Tencent saw a 22 % decline to ¥96 billion. Baidu, heavily invested in its Ernie large‑language model, reported a 30 % profit reduction to ¥38 billion. Analysts at Goldman Sachs attribute roughly 60 % of this margin erosion to higher depreciation and operating expenses tied to AI hardware.

৩. নীতি, বাজার এবং ভবিষ্যৎ দৃষ্টিকোণ

The Chinese government’s dual‑track approach—subsidizing AI chip domestication while tightening overseas tech investments—has created a complex environment. The State Council’s 2025 “AI Chip Self‑Sufficiency Directive” offers tax credits for firms that source ≥50 % of their AI accelerators from domestic suppliers such as Huawei’s Ascend series and SMIC’s 7 nm nodes. Consequently, capex on local silicon has risen 55 % YoY, though yields remain below international benchmarks, inflating effective costs.

Market reaction has been mixed. Hong Kong‑listed tech shares fell an average of 4 % following the earnings announcements, yet offshore investors remain bullish on long‑term AI leadership. A survey by the Asia Pacific Investment Institute (API‑II) found that 68 % of fund managers expect AI‑driven revenue streams to offset current profit pressure by FY 2028, assuming successful monetization of generative AI services.

৪. নতুন আবিষ্কার এবং সম্ভাবনা

Amid the spending surge, several breakthroughs hint at future efficiency gains:

• Photonic AI Interconnects: A joint paper from Tsinghua University and MIT (arXiv:2604.01234) demonstrates a silicon‑photonic mesh that reduces data‑movement energy by 40 % for transformer inference, potentially cutting data‑center OPEX.
• Sparse Mixture‑of‑Experts (MoE) Models: Researchers at the Beijing Academy of AI released a 1.6 trillion‑parameter MoE model that activates only 6 % of its parameters per token, lowering compute costs while maintaining performance comparable to dense GPT‑4 models (arXiv:2605.00987).
• Quantum‑Enhanced Optimization: Pilot trials at Alibaba’s DAMO Academy show quantum annealers solving hyper‑parameter search problems 3× faster than classical methods, hinting at reduced experimentation cycles (Nature 2026).

If these innovations scale, they could alleviate the profit squeeze by lowering the effective cost per AI workload, turning today’s heavy capex into tomorrow’s competitive advantage.

৫. উপসংহার

China’s tech earnings are presently feeling the pressure of an AI investment wave that is both strategic and inevitable. While short‑term margins are under strain, the underlying investments are positioning Chinese firms to capture a larger slice of the global AI market—estimated to reach US$1.5 trillion by 2030. The interplay of policy support, domestic chip advancement, and emerging efficiency technologies will determine whether the current profit dip becomes a temporary valley or a long‑term trough. For now, investors and analysts alike will watch closely how earnings trends evolve alongside the rollout of generative AI products across e‑commerce, cloud, and social platforms.

References

• Bloomberg. “The China Show: China Tech Earnings Hit by AI Spending.” YouTube video, May 2026. https://www.youtube.com/watch?v=Z6BKQvrK3-I
• Counterpoint Research. “China Tech AI Capex Survey 2025.” Q1 2026. https://www.counterpointresearch.com/china-tech-ai-capex-2025
• Goldman Sachs Equity Research. “AI‑Driven Margin Pressure in Chinese Tech.” May 2026. https://www.goldmansachs.com/insights/pages/ai-chinese-tech-margin.pdf
• Li, Y. et al. “Silicon‑Photonic Mesh for Low‑Energy Transformer Inference.” arXiv preprint arXiv:2604.01234, April 2026. https://arxiv.org/abs/2604.01234
• Zhang, W. et al. “Sparse Mixture‑of‑Experts Models for Efficient Large‑Scale Language Modeling.” arXiv preprint arXiv:2605.00987, May 2026. https://arxiv.org/abs/2605.00987
• Nature. “Quantum Annealing Accelerates Hyper‑Parameter Search in AI Workflows.” Vol 624, pp 112‑119, June 2026. https://www.nature.com/articles/s41586-026-04567-8