上海— 中国科技巨头华为周一宣传了一种开发先进半导体的新方法，尽管美国制裁,，因为英伟达很难在中国销售其高端芯片。
SHANGHAI — Chinese tech giant Huawei on Monday touted a new approach to developing advanced semiconductors despite U.S. sanctions, as Nvidia struggles to sell its high-end chips in China.
华为表示，今年秋天，它开发了一种名为"LogicFolding"的新工程方法来制造其麒麟智能手机芯片。
Huawei said it developed a new engineering approach called "LogicFolding" to manufacture its Kirin smartphone chips this fall.
That breakthrough comes as Nvidia faces U.S. export restrictions in China and Apple contends with renewed competition from Huawei in the world的 second-largest consumer economy.
华为的 Mate 60 智能手机, 于 2023 年推出, 包含由先进芯片提供支持的 5G 连接，帮助该公司从苹果手中夺回了市场份额。
Huawei的 Mate 60 smartphone, launched in 2023, included 5G connectivity powered by an advanced chip that helped the company regain market share from Apple.
尽管近年来美国的限制阻止英伟达向中国出售其最先进的芯片,，但北京方面却大力支持本土技术。上周, Nvidia 首席执行官黄仁勋 (Jensen Huang) 告诉 CNBC，这家美国芯片制造商已 " 将中国市场让给了华为。
While U.S. restrictions have kept Nvidia from selling its most advanced chips to China in recent years, Beijing has pushed to support homegrown technology instead. Last week, Nvidia CEO Jensen Huang told CNBC the U.S. chipmaker had "conceded" the Chinese market to Huawei.
"对于 Nvidia, 来说，这意味着向中国销售 H200 等先进芯片的窗口正在缩小," 乔治·陈, 合伙人兼亚洲集团数字业务联席主席表示。
"For Nvidia, this means the window to sell advanced chips such as the H200 into China is narrowing," said George Chen, partner and co-chair of digital practice at The Asia Group.
"这一轨迹可能会加剧华盛顿的担忧,，华为仍然是美国出口限制的象征," 他说。
"This trajectory will likely heighten concerns in Washington, where Huawei remains emblematic of U.S. export restrictions," he said.
华为表示，到2031,，其新芯片技术可提供相当于1.4纳米工艺技术—的能力，而全球芯片领导者台积电已开始量产2纳米芯片。
Huawei said that by 2031, its new chip technology could deliver capabilities equivalent to 1.4-nanometer process technology — while global chip leader TSMC has begun volume production of 2-nanometer chips.
Nanometer processes refer to chip manufacturing technology, with smaller nodes typically enabling faster and more efficient semiconductors.
DGA Group, 技术, 亚洲和美洲, 技术主管 Paul Triolo, 对华为的 1.4 纳米说法持怀疑态度。
Paul Triolo, head of technology, Asia and Americas, at DGA Group, was skeptical of Huawei的 1.4-nanometer claim.
"A stacked/folded design can produce effective density gains, but it does not mean Huawei has solved the full process, yield, power, thermal, and device-performance problems associated with true 1.4 nm-class manufacturing," he said.
Counterpoint Research 研究副总裁 Neil Shah, 表示，由于华为无法评估荷兰芯片设备制造商 ASML, 的先进极紫外, 或 EUV, 光刻机，华为被迫寻求芯片开发的替代方案，以保持在 AI, 领域的竞争力。
Blocked from assessing advanced extreme ultraviolet, or EUV, lithography machines from Dutch chip equipment maker ASML, Huawei has been forced to pursue alternatives to chip development as it seeks to remain competitive in AI, said Neil Shah, vice president of research at Counterpoint Research.
"However, this parallel semiconductor path is still unproven at scale. This approach can introduce tough thermal constraints and packaging complexities that can hit the manufacturing yields," Shah said.
Huawei的 efforts to deploy the technology in its flagship Mate 90 smartphone series this fall would mark an engineering feat, but scaling it to AI datacenters would serve as the "ultimate litmus test for China的 creative workaround to Western sanctions," he added.
Academic ambitions
华为还在为其半导体研究寻求更大的学术认可。 On Monday, the company described its findings as the "Law of Tau," or "τ scaling," and claimed it addresses challenges faced by the semiconductor industry.
Huawei is also seeking greater academic recognition for its semiconductor research. On Monday, the company described its findings as the "Law of Tau," or "τ scaling," and claimed it addresses challenges faced by the semiconductor industry.
Huawei said it has designed and mass-produced 381 chips based on the "τ Scaling Law" over the last six years.
Semiconductor development has, for decades, relied on "Moore的 Law," an observation that the number of transistors would double roughly every two years — delivering more computing power while lowering costs.然而,甚至Nvidia的 Huang也表示摩尔的定律不再适用于未来的芯片开发。
Semiconductor development has, for decades, relied on "Moore的 Law," an observation that the number of transistors would double roughly every two years — delivering more computing power while lowering costs. However, even Nvidia的 Huang has said Moore的 Law is no longer applicable to future chip development.
"华为正在将工程战略转变为准'law,'" Triolo 说。
"Huawei is turning an engineering strategy into a quasi-'law,'" Triolo said.
新原则"更多的是系统级优化原则:缩短线路,堆栈逻辑,改进内存语义,和协同设计芯片,封装,软件,和集群,"他说。
The new principle "is more a systems-level optimization doctrine: shorten wires, stack logic, improve memory semantics, and co-design chips, packages, software, and clusters," he said.
, Triolo 表示，围绕热管理和大规模制造的挑战仍然存在。, Triolo 表示。
Still, challenges remain around heat management and manufacturing at scale, Triolo said.
Huawei的 new chip architecture expands the layout from one layer to two, significantly increasing power efficiency, according to Tingbo He, president of Huawei的 semiconductor business.
这种结构可以让晶体管在更多点上相互交互, 同时担任公司的 科学家委员会主任, 的他, 在电气和电子工程师协会 国际电路与系统研讨会上表示。
This structure allows transistors to interact with each other at more points, He, who is also a director of the company的 scientist committee, said at the Institute of Electrical and Electronics Engineers International Symposium on Circuits and Systems.
However, she acknowledged that challenges remain, as Huawei is only just beginning a decade-long development path for the new technology.
News Tips
收到机密新闻提示? 我们希望收到您的来信。
CNBC 时事通讯
CNBC Newsletters
注册免费时事通讯，将更多 CNBC 新闻发送到您的收件箱
将此信息发送到您的收件箱,，并获取有关我们产品和服务的更多信息。
与我们一起做广告
Advertise With Us
© 2026 Versant Media, LLC。版权所有。 Versant 媒体公司。
数据为实时快照 *数据至少有15分钟延迟。全球商业和金融新闻, 股票行情, 以及市场数据和分析。