量子科技在 2025 年加速走出实验室并走向地缘竞赛;联合国将 2025 定为「国际量子科学与技术年」。它以量子位元(qubit)的叠加等特性追求指数级效能提升;麦肯锡估计到 2035 年可带来最高约 970 亿美元的全球营收,贝恩则估算整体生态系可解锁最高约 2500 亿美元的市场价值,资金与关注正快速涌入。
2025 年多项里程碑密集出现:微软 2 月发布首款量子晶片并宣称有路径在单一处理器上容纳 100 万 qubit;亚马逊云端团队称 Ocelot 可将量子纠错成本最多降低 90%;IBM 6 月提出到 2029 年启动「容错」量子电脑的细致框架;Google 10 月表示在 Willow 上跑出可在其他量子系统重复验证的「Quantum Echoes」,并称其速度较最强超级电脑快 13,000 倍。
美国仍领先,但中国快速逼近:北京已为量子运算投入 153 亿美元公共资金,美国承诺 19 亿美元,前者超过后者 8 倍。产业时间表分歧(例如「15–30 年」到「未来几年」即可解决部分问题,甚至激进预估也多指向至少 5 年后才有明显效用),反映 qubit 稳定与降错的难度;但资安风险更迫切:贝恩调查显示 73% 资安专业人士预期 5 年内成为重大风险(32% 认为 3 年内),然而仅 9% 已有应对计划。
Quantum tech moved from lab curiosity to geopolitical and commercial urgency in 2025, the UN’s International Year of Quantum Science and Technology. With qubits enabling exponential-style gains, McKinsey projects up to $97B in global revenue by 2035, while Bain estimates up to $250B in market value across the broader ecosystem. Capital and attention are rising quickly as timelines feel compressed compared with the long-running “always a decade away” narrative.
Milestones piled up fast: Microsoft (Feb) promoted a path to 1M qubits on a single processor; Amazon’s Ocelot claimed up to 90% lower quantum error-correction cost; IBM (Jun) published a detailed fault-tolerant roadmap targeting 2029; Google (Oct) said its Willow chip ran a “verifiable” algorithm, “Quantum Echoes,” 13,000× faster than the world’s top supercomputer. The pace is dramatically higher than even five years ago, though hype-versus-science concerns persist.
The US still leads, but China is narrowing the gap: Beijing has earmarked $15.3B in public quantum funds versus $1.9B pledged by the US—over 8×. Forecasts also diverge (15–30 years to broad usefulness vs “interesting problems” in the coming years; even aggressive views put meaningful utility at least 5 years out), reflecting qubit stability and error-rate challenges. Security urgency is clearer: a Bain survey found 73% expect post-quantum cryptography to be a material risk within 5 years (32% within 3), yet only 9% say they have a plan.