Nvidia's former No. 3 Jay Puri drawn into $2.5 billion China chip scandal
The link highlighted by Muddy Waters was a set of publicly available photographs showing former Nvidia executive Jay Puri with a woman identified as Cenly Chen or Cenly Puri. The firm said the images raised a question because a former Super Micro executive named Cenly Chen was involved in Nvidia chip purchases and distribution. That connection matters because Super Micro faces allegations of diverting chips to China. The report pointed to images from the San Francisco Opera’s Masked Ball in 2024 and events on April 23, 2026. It compared them with a photograph of Super Micro’s Cenly Chen at Nvidia’s GPU Technology Conference. Muddy Waters sought clarification from the investor relations firm representing Chronoscale, where Chen is now chief executive. The firm received no response. The photographs show the two appearing together at public events, but the report’s central question remains unresolved. They do not establish that the woman with Puri is the same Cenly Chen from Super Micro. Nor do they prove that either person took part in wrongdoing. The article describes the connection as a question raised for scrutiny, not a confirmed finding.
What evidence links Jay Puri to Cenly Chen, and what does that evidence establish—or leave unproven?
The link highlighted by Muddy Waters was a set of publicly available photographs showing former Nvidia executive Jay Puri with a woman identified as Cenly Chen or Cenly Puri. The firm said the images raised a question because a former Super Micro executive named Cenly Chen was involved in Nvidia chip purchases and distribution. That connection matters because Super Micro faces allegations of diverting chips to China.
The report pointed to images from the San Francisco Opera’s Masked Ball in 2024 and events on April 23, 2026. It compared them with a photograph of Super Micro’s Cenly Chen at Nvidia’s GPU Technology Conference. Muddy Waters sought clarification from the investor relations firm representing Chronoscale, where Chen is now chief executive. The firm received no response.
The photographs show the two appearing together at public events, but the report’s central question remains unresolved. They do not establish that the woman with Puri is the same Cenly Chen from Super Micro. Nor do they prove that either person took part in wrongdoing. The article describes the connection as a question raised for scrutiny, not a confirmed finding.
What role did the shareholder lawsuit allege Chen played in buying Nvidia GPUs and deciding where they went?
A shareholder lawsuit alleged that Cenly Chen had a central role in Super Micro’s Nvidia GPU business. As chief growth officer, she allegedly helped manage the company’s acquisition of GPUs and decide how its limited supply was distributed among customers. These claims matter because the lawsuit tied her work to an alleged scheme to sell chips to China.
The complaint alleged that Chen had a close relationship with Nvidia and met regularly with Super Micro co-founder and executive Liang to discuss Nvidia dealings and get approval for allocations. An account attributed to a former employee said Chen requested that Nvidia release chips carrying serial numbers associated with GPUs allegedly diverted to Chinese customers. These are claims in the complaint, not proven findings.
The article says Chen left Super Micro in May 2026, after public disclosure of the alleged export-diversion scheme. The complaint alleged she resigned so the company could argue that no current employees were involved. It also asserted that she could not have been unaware of the alleged misconduct. The source presents these points as allegations in a civil lawsuit, not established criminal wrongdoing.
How did prosecutors allege the servers containing Nvidia chips were routed to Chinese buyers?
The supplied text does not describe a route used to move servers containing Nvidia chips to Chinese buyers. It names no intermediaries, shipping locations, paperwork, or transfer method. The article’s visible discussion establishes only that a shareholder complaint alleged Nvidia GPUs were diverted to China and that the alleged scheme involved Super Micro.
That distinction matters: a claim that chips were diverted does not, by itself, explain how servers were transported or who handled each step. The excerpt ends just as it begins a section titled “$2.5 BILLION CHIP DIVERSION SCHEME.” It therefore does not provide the details needed to answer the question about prosecutors’ allegations. No routing account can be reconstructed from the text provided.
The article also describes the allegations against Chen as claims in a civil shareholder lawsuit, not established findings of criminal wrongdoing. It does not identify prosecutors or give a prosecution account of server movements in the portion supplied. More detail may exist beyond the excerpt, but it is not present here, so a specific route or mechanism cannot be stated without adding unsupported information.
How much equipment was allegedly involved in the overall scheme, and how much was reportedly diverted in a short period in 2025?
The article excerpt does not provide the equipment totals asked for. It refers to a shareholder lawsuit over the alleged smuggling of USD 2.5 billion worth of chips to China, but it does not say how many servers or other pieces of equipment were involved overall. It also gives no figure for a shorter diversion period in 2025.
Those distinctions matter because a dollar value for chips is not the same as a count of servers or a time-bounded quantity. The supplied text does not include the section that might set out further details: it stops after the heading “$2.5 BILLION CHIP DIVERSION SCHEME” and the opening words of that section. It would therefore be inaccurate to supply equipment counts or 2025 totals from this excerpt.
What the text does say is that a consolidated shareholder class-action complaint filed against Super Micro in October 2026 alleged Nvidia GPU diversion to China. It frames the claims as allegations, not established findings of criminal wrongdoing. The only scale figure visible here is USD 2.5 billion worth of chips; the other requested figures are not stated.
What does it mean for a chip or server to be export-controlled?
In general, an export-controlled chip or server is an item subject to government rules limiting its export, re-export, or transfer to certain destinations, users, or uses. Controls may cover a product because of its technical capabilities or its intended recipient or end use. The rules and licensing requirements depend on the jurisdiction and item.
For example, a company may need government authorization before sending a controlled technology item to a particular country or customer. Transferring it through another country or company may also be regulated, depending on the applicable rules. A product being advanced does not, by itself, tell us which controls apply; its classification and the circumstances of the transaction matter.
The supplied article alleges that Nvidia GPUs were diverted to China and describes claims about Super Micro’s chip allocations. It does not identify the specific export-control classification, rule, or license requirement for the chips or servers involved. This definition is general background, not a finding about the alleged transactions. The legal status of any particular shipment would depend on facts and rules not provided in the excerpt.
What consequences can follow when companies divert export-controlled technology to buyers who are not allowed to receive it?
In general, companies and individuals that violate export-control rules can face consequences such as civil or criminal penalties, depending on the law, conduct, and evidence. Authorities may also restrict a company’s ability to export or require it to strengthen compliance controls. Buyers, shipments, and related business can face added scrutiny. These are possible consequences under export-control regimes, not outcomes stated for this case.
For instance, if a shipment required a license and was sent to a prohibited recipient without one, investigators could examine who arranged the transfer and what they knew. Responsibility and penalties would depend on the applicable rules and proven facts. A claim in a lawsuit alone does not establish a violation or determine what consequences follow.
The article describes allegations that Nvidia chips were diverted to China and says the claims against Chen remain allegations in a civil lawsuit, not established findings of criminal wrongdoing. It does not report specific penalties or legal outcomes. Any consequences in this matter would depend on further evidence and legal proceedings; the excerpt does not say what will happen next.
What is a GPU, and why are powerful GPUs important for developing and running artificial intelligence?
A GPU, or graphics processing unit, is a processor designed to carry out many calculations in parallel. It was originally developed for graphics work, which often involves performing similar calculations across many pixels. GPUs can also be used for other computing tasks that benefit from parallel processing.
Developing many AI systems involves doing large numbers of mathematical operations on data. A powerful GPU can process many of these operations at the same time, making it useful for training an AI model. Once a model is trained, GPUs can also help run it on new inputs, a process often called inference. The exact hardware needs vary by model and task.
The article identifies Nvidia GPUs as the chips at the center of the alleged diversion and says Chen’s role involved their purchase and distribution. It does not explain GPU architecture or AI computing. This is general background: it helps explain why high-performance processors matter, but it does not establish what specific AI work the GPUs in the lawsuit were intended to support.
This brief was written by AI from the original reporting and checked by other models. Names, figures and quotes come from the source; read it for full context.
Read more in the JupiteX app
Pulse is free. New stories every 4 hours, each one broken into the questions that explain it.
Or read more news on the web