China Aims For Twin-Track Use Of Nvidia H200 Amid Back Door Fears

Nearly a decade later, as the Great Leap Forward began in 1958, Mao promoted the concept of“walking on two legs,” calling for the full mobilization of all positive forces and warning against one‐sided or unbalanced development. In practical terms, the slogan urged China to advance heavy industry alongside manufacturing and agriculture, and to pursue production growth while also safeguarding people's livelihoods.

A Chinese columnist writing under the pen name Tangyipao gives a detailed explanation of the“two-track” approach, highlighting the H200's strategic value for China.

“First, we need to understand what the H200 really offers,” he writes.“The H200 is Nvidia's second-tier flagship chip, just below its latest Blackwell series, with its core advantage lying in high-bandwidth memory that significantly boosts data throughput. The H200 delivers nearly six times the performance of the H20.”

“For AI research, high-end AI chips are essentially engines of computing power. Access to the H200 will enable Chinese laboratories to build systems that approach top-tier US performance for large-language model (LLM) training and complex data processing,” he says. Famous LLMs include ChatGPT and DeepSeek.

“While benefiting from the relaxation of the H20 export, we must stay rational,” Tangyipao says.“China can make reasonable use of advanced foreign technology to promote industrial development, but it must also remain alert to the trap of technological lock-in. After all, the US government's 25% cut in Nvidia's H200 revenue from China is actually a tax on Chinese buyers. China needs to press ahead with domestic chip development.”

“With the H200, China can accelerate technological iteration and better pursue a 'two-track' approach. Even without it, China's AI industry can continue to move forward on a path of independent development,” he says.“As domestic processors such as Huawei's Ascend mature, China's reliance on foreign high-end chips will gradually decline.”

Mao Yu, a columnist at the Shanghai-based IT Times, has this to say:“Compared with the degraded H20, the H200 offers a significant performance jump. But its real appeal is not just hardware. It is Nvidia's deeply entrenched CUDA ecosystem.” Citing an unnamed industry expert, Mao says several major Chinese technology firms are preparing to order close to 100,000 H200 chips.

“The reason big Chinese companies are willing to buy is that the H200 is essentially plug-and-play,” he says.“After more than a decade of development, the CUDA architecture has formed a complete closed loop, from development tools to application scenarios. It has become the infrastructure for AI training, an advantage domestic chips cannot replicate in the short term.”

He says people should not overworry that the H200 will dominate the Chinese market.

“Only a handful of large Chinese companies are engaged in LLM training, and many of them have already stockpiled key resources such as HBM memory chips, with the H200 to be serving mainly as a supplement to existing clusters,” he says.“Most small and medium-sized enterprises focus on inference workloads with limited need for cutting-edge computing power, while government-backed cloud providers and users in specific sectors continue to favor domestic AI chips.”

“While domestic chip makers still need some time to catch up to Nvidia in terms of high-end AI chip designs and production capacity, the H200 will help China ease its computing power bottlenecks and improve research efficiency,” he says.

Reuters also reported that Chinese technology firms, including ByteDance and Alibaba, are keen to purchase the H200 chips.

The Trump administration approve d exports of Nvidia's H20 chips to China in July, but the decision was quickly met with skepticism from Chinese commentators. Many dismissed the chips as“poisoned wine,” arguing that even with reduced performance, they would further bind domestic AI firms to Nvidia's technology stack and slow efforts to build an independent ecosystem.

Beijing subsequently raised concerns over alleged“back doors” in the H20, with commentators warning that such features could allow Nvidia or US authorities to locate, monitor or even remotely control GPUs deployed overseas.

Nvidia has repeatedly rejected those allegations. In a blog post published on December 10, the company said it was developing a software-based monitoring tool designed to help customers manage large GPU deployments, not to exert control over them.

Sign up for one of our free newsletters

• The Daily Report Start your day right with Asia Times' top stories
• AT Weekly Report A weekly roundup of Asia Times' most-read stories

“Nvidia is developing a software solution for visualizing and monitoring fleets of Nvidia GPUs, giving cloud partners and enterprises an insights dashboard that can help them boost GPU uptime across computing infrastructures,” the company said.

Nvidia emphasized the service would be optional and software-driven.

“This optional service provides real-time monitoring by each GPU system communicating and sharing GPU metrics with the external cloud service,” the company said, adding that“Nvidia GPUs do not have hardware tracking technology, kill switches and back doors.”

According to Nvidia, the service will feature a client software agent that the customer can install to stream node-level GPU telemetry data to a portal hosted on Nvidia GPU Cloud (NGC).

In other words, Nvidia and its clients will be able to locate all the company's new chips overseas.

Coincidentally, The Information reported on the same day that Chinese artificial intelligence startup DeepSeek was using smuggled Nvidia's Blackwell chips to build“phantom data centers” to develop new AI models in China. The US has banned the shipment of Blackwell chips to China.

Responding to the report, an Nvidia spokesperson said the company had seen no evidence of large-scale chip smuggling. The spokesperson said Nvidia had not received credible information suggesting that data centers were built to deceive the company or its partners and then secretly dismantled and moved elsewhere, adding that such scenarios were unlikely.

Some observers said Beijing is less concerned about Nvidia's ability to track the physical location of its chips than about whether the company's assurances of having“no kill switch” can be independently verified.

Jin Mei, a Guangdong-based writer, says any potential back door functionality must be clearly verified, not merely addressed through Nvidia's assurances. She says Chinese buyers must seek case-by-case regulatory approval and explain why Nvidia chips are needed instead of domestic alternatives.

Read: AMD's MI308 chips may face H20‐like headwinds in China

Follow Jeff Pao on Twitter at @jeffpao3

Sign up here to comment on Asia Times stories Or Sign in to an existing accoun

Thank you for registering!

An account was already registered with this email. Please check your inbox for an authentication link.

• Click to share on X (Opens in new window)
• Click to share on LinkedIn (Opens in new window) LinkedI
• Click to share on Facebook (Opens in new window) Faceboo
• Click to share on WhatsApp (Opens in new window) WhatsAp
• Click to share on Reddit (Opens in new window) Reddi
• Click to email a link to a friend (Opens in new window) Emai
• Click to print (Opens in new window) Prin