Free tools Windows power users keep installed
One-click scans. No signup required.
Reuters reported on July 28, 2025, citing two unnamed sources, that Nvidia had placed an order with TSMC for 300,000 H20 AI chips. The report described an order—not proof that the chips were manufactured, exported, delivered or sold. It also said the export licenses needed for sales to China had not been approved at that time.
What did Nvidia reportedly order from TSMC?
The reported 300,000-chip order was intended to increase Nvidia’s H20 supply amid strong demand in China, Reuters said. Reuters’ sources estimated that Nvidia already held 600,000 to 700,000 H20 chips in inventory. Those figures were reported estimates, not audited totals.
Reuters also reported that Nvidia had sold around one million H20 chips in 2024, attributing the figure to research firm SemiAnalysis. The inventory and sales estimates describe different things and should not be treated as a verified count of chips available for shipment to China.
| Figure | What it describes | Source and qualification |
|---|---|---|
| 300,000 chips | Reported new order to TSMC | Reuters, 2025; based on two unnamed sources |
| 600,000–700,000 chips | Reported Nvidia H20 inventory at the time | Reuters, 2025; estimate from unnamed sources |
| Around 1 million chips | H20 chips Nvidia reportedly sold in 2024 | SemiAnalysis figure attributed by Reuters, 2025 |
Could Nvidia sell H20 chips to China again?
In the July 2025 reporting, a change in policy did not mean that shipments were already cleared. The sequence involved several distinct steps:
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problems#1 Best Overall
- PLEASE NOTE: Exporting an NVIDIA RTX Pro 6000 GPU outside the US requires strict adherence to the U.S. Export Administration Regulations (EAR) and issuance of an export license from the Bureau of Industry and Security (BIS). Compliance and Know Your Customer (KYC) screening may be required as a condition of order acceptance. [NVIDIA Blackwell Streaming Multiprocessor] The new SM features increased processing throughput, and new neural shaders that integrate neural networks inside of programmable shaders | DLSS 4: Multi Frame Generation ensures ultra-smooth frame pacing for lifelike simulations.
- [Double-Flow-Through Design] The RTX PRO 6000 Blackwell features a double-flow-through cooling design, optimizing efficiency and airflow to sustain peak performance under 600W power loads. | [5th Gen Tensor Cores] Deliver up to 3X the performance of the previous generation and support for FP4 precision for faster AI model processing times with reduced memory usage, enabling local fine-tuning of LLMs and generative AI | [4th Gen Ray Tracing Cores] Double the ray-triangle intersection rate of the previous generation to create photoreal, physically accurate scenes and immersive 3D designs with RTX Mega Geometry, which enables up to 100X more ray-traced triangles.
- [PCIe Gen 5] Support for PCIe Gen 5 provides double the bandwidth of PCIe Gen 4, improving data-transfer speeds from CPU memory and unlocking faster performance for data-intensive tasks like AI, data science, and 3D modeling. | [GDDR7 Memory] With 96 GB of GPU memory and 1.8 TB ps bandwidth, it can tackle massive 3D and AI projects, fine-tune AI models locally, explore large-scale VR environments, and drive larger multi-app workflows.
- [DisplayPort 2.1] Achieve unparalleled visual clarity and performance, driving high resolution displays at up to 8K at 240 Hz and 16K at 60 Hz. Increased bandwidth enables seamless multi-monitor setups while HDR and higher color depth support ensures superior color accuracy for precision work, such as video editing, 3D design, and live broadcasting.
- [Universal MIG] Divide a single RTX PRO 6000 Blackwell into multiple isolated instances, each with dedicated resources, allowing for concurrent execution of multiple workloads, optimized GPU utilization, and secure isolation of different applications or users. [WARRANTY] 3 YR Manufacturer's Warranty. Bulk OEM Packaging. Retail Packaging is NOT included.
- Late 2023: After U.S. restrictions on Nvidia’s other AI chipsets, Nvidia designed H20 for the Chinese market.
- April 2025: U.S. controls effectively barred H20 sales to China.
- July 2025: The Trump administration said sales would be allowed to resume. Nvidia said it was applying for licenses and had received assurances they would be granted, as reported by Axios on July 15.
- July 28, 2025: When Reuters reported the new TSMC order, it said the required licenses had not yet been approved.
An assurance that licenses would be granted, an issued license, permission to export a particular shipment, and delivery to a customer are not interchangeable. The cited July 2025 coverage does not establish whether the reported order was ultimately manufactured, delivered or sold, or what the license status was on October 4, 2026.
Why was there demand for Nvidia’s H20 in China?
H20 was a China-focused product made in response to earlier U.S. export restrictions. Reuters described it as less powerful than Nvidia’s H100 and Blackwell-series chips sold outside China; the reporting did not provide a comparable benchmark or full specification table.
H20’s role was not simply a matter of peak computing power. Nvidia CEO Jensen Huang emphasized its memory bandwidth for AI workloads. At an event in Beijing, as reported by Reuters on July 15–16, 2025, he said: “H20 was released from its ban, the memory bandwidth is extremely good, for LLMs and other new models it will be excellent.”
Rank #2
- NVIDIA Volta GV100 Architecture — 4,608 CUDA Cores, 640 1st-Gen Tensor Cores delivering 14 TFLOPS FP32 and 112 TFLOPS deep learning performance for AI training, inference, HPC, and scientific computing workloads
- 32GB HBM2 ECC Memory — 900 GB/s Bandwidth — High-bandwidth memory on a 4096-bit bus with ECC error correction provides the memory capacity and throughput required for the largest AI models, simulations, and datasets
- PCIe 3.0 x16 Interface — 250W TDP — Standard PCIe Gen3 connectivity with passive cooling designed for enterprise rack server deployment in HPE ProLiant, Dell PowerEdge, and Supermicro platforms with adequate chassis airflow
- NVLink — Scale to 96GB Unified Memory — Connect two V100 GPUs via NVLink at 300 GB/s bi-directional bandwidth to scale GPU memory from 32GB to 96GB for larger AI training and HPC workloads
- Multi-Precision Computing — Supports FP64 (7 TFLOPS), FP32 (14 TFLOPS), FP16 (112 TFLOPS) and INT8 precision modes for flexible deployment across training, inference, and scientific simulation workloads
Reuters reported that Chinese technology companies including Tencent, ByteDance and Alibaba had substantially increased H20 orders before the April 2025 ban while deploying DeepSeek’s models and their own. That is context for demand, not evidence that any of those companies bought chips from this specific 300,000-unit order. Reuters also described Huawei as a rival whose chips were viewed as less powerful in that report, but did not establish that as an independent benchmark comparison.
Recommended Free Tools
How long would restarting supply take?
Reuters reported that Huang said restarting production would take about nine months. That was a contemporaneous estimate for restarting the supply chain, not a promised delivery date for the 300,000-chip batch. The Associated Press likewise reported on July 16, 2025, that restoring production would take time and that recovery depended on order volume and how quickly Nvidia could meet demand.
The policy disruption also carried a cost at the time. Reuters reported $5.5 billion in inventory write-down exposure following the April ban and cited Huang’s estimate of $15 billion in potential sales forgone, which came from his comments on the Stratechery podcast. These were period-specific exposure and potential-sales figures, not a statement of Nvidia’s present financial impact.
Rank #3
- Professional GPU with Blackwell Architecture
- Blackwell Architecture
- 24GB GDDR7 with PCIe 5.0 & Ray Tracing
- AI Workstation
What Huang said about future China sales
Reuters quoted Huang saying: “I hope to get more advanced chips into China. Today H20 is still incredibly good, but in coming years, whatever we are allowed to sell to China we will do so.” In an Associated Press account on July 16, 2025, Huang described Nvidia’s influence over the policy decision this way: “We can only influence them, inform them, do our best to provide them with facts,” adding, “And then beyond that is out of our control.”
Those remarks convey Nvidia’s position and intentions; they do not establish regulatory approval or a later shipment. The 300,000-chip figure remains a reported order in Reuters’ July 2025 account, rather than confirmation that the batch reached customers.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
