TMTPost reports that Duan Yongping reduced his Nvidia common-stock holdings from 13.8438 million shares in 2026 Q1 to 6.281 million in Q2 and 2.319 million in Q3. Duan said he remains positive on Nvidia's business but will not add to the position while waiting to evaluate the permanent CEO, and that he may buy back shares previously called away. The same article was republished by 创业邦科技, so it is not an independent corroboration.
CNBC reports that Nvidia interim CEO Colette Kress told “Mad Money” that the chipmaker's customers and employees remain committed to the company. The report does not provide independent evidence or further details about the basis for her assertion.
Jeff Bezos advised NVIDIA’s board to prioritize an internal candidate if Jensen Huang did not return as CEO. Speaking at Semafor’s Silicon Valley & The World conference, Bezos said NVIDIA’s complexity makes an external executive less likely to succeed and described outside hiring as a last resort. NVIDIA’s board has said it will consider both internal and external candidates.
According to The Information, former Nvidia sales chief Jay Puri oversaw major chip customers, including Alice Huang, and controlled sales to Supermicro. Puri is married to a former Supermicro executive. The Information also reported that Huang worked with a Chinese firm accused of secretly importing at least $7.5 billion in restricted Nvidia AI chips into China, circumventing U.S. export restrictions; the allegations were not independently established by the supplied reports.
Nvidia’s first earnings update since longtime leader Jensen Huang’s departure beat expectations, raised its growth outlook and issued a next-year sales forecast above Wall Street estimates, according to Bloomberg Technology. The company also pledged to improve access to critical supplies and added $150 billion to its stock-buyback program. Nvidia sought to damp speculation that Huang might return, while the CEO search and talent risks remain unresolved questions.
A report says Nvidia is no longer shipping Hopper data-center chips to China, despite receiving US government approval last year to sell a version of the model in the Chinese market.
Tomasz Tunguz says NVIDIA faces a choice about which customers to support as it signals tighter financing terms for neoclouds. Non-hyperscalers’ share of NVIDIA Data Center revenue slipped from 45% to 44% in Q3, while AI Clouds, Industrial & Enterprise revenue reached $47.5 billion, according to NVIDIA disclosures cited in the analysis.
Nvidia disclosed up to $118 billion in potential guarantee exposure to partners, including up to $105 billion in credit support for OpenAI-related data center construction. The figures represent maximum contingent exposure rather than completed spending or confirmed debt, and the arrangements will take effect in stages as projects are built; as of October 25, Nvidia had made no payments under the guarantees.
NVIDIA interim CEO and CFO Colette Kress raised the company’s forecast for next-fiscal-year revenue growth from about 70% to 75%, citing expanding demand and additional secured supply, TechNews reported. She also said the fourth quarter was a relative low point for gross margin, maintained confidence in the fiscal 2028 margin outlook, and said the company had assessed new U.S. export controls but still needed further study of their challenges.
Nvidia interim CEO Kress said that under multiyear agreements, Hopper GPU re-contracting prices can reach twice the original price. The report also said the B300's price has increased by more than 60% since launch.
NVIDIA interim CEO and CFO Colette Kress said the cloud computing industry's order backlog had reached $2.5 trillion, up from $2 trillion three months earlier, according to 36氪快讯. She also said revenue growth from NVIDIA's hyperscale cloud customers had accelerated for six consecutive quarters. Kress attributed growth in AI applications and related revenue to new-generation frontier models from OpenAI and Anthropic.
AWS Machine Learning Blog says Amazon SageMaker HyperPod Inference now supports NVIDIA Dynamo for multi-node model serving, disaggregated prefill and decode, and KV-aware routing. The managed add-on installs Dynamo components and provides endpoint configuration, GPU scheduling, node recovery, and autoscaling, with deployment examples using GLM-5.3 on vLLM.
Reuters, as relayed by IT之家, reports that Jensen Huang remains eligible under Nvidia's proxy-access rules to nominate up to two independent director candidates after resigning as CEO and a director. LSEG data cited by IT之家 shows Huang and affiliated entities hold about 3.5% of Nvidia voting power for more than three years, meeting the eligibility threshold. Nominees would still need to satisfy Nvidia's independence and governance requirements and receive shareholder support, while Huang and Nvidia have not said whether he plans to exercise the right.
36Kr reported that NVIDIA’s shares rose 5% in after-hours trading. The company forecast fourth-quarter revenue of $126.1 billion to $133.9 billion, above market expectations, while its board approved an additional $150 billion stock-buyback plan.
36Kr reported that NVIDIA generated $115.8 billion in fiscal 2027 third-quarter revenue, up 103% year over year, and net income of $65.3 billion, up 105%. GAAP and non-GAAP gross margins were 74.8% and 74.9%, while diluted EPS was $2.70 and $2.68, respectively. The company returned $28.1 billion to shareholders through buybacks and cash dividends during the quarter. 36Kr also reported that NVIDIA’s board approved an additional $150 billion in share-repurchase authorization in September 2026, with no expiration date.
A top Microsoft executive said Nvidia’s corporate drama could prompt customers to consider alternatives to the company’s AI chips. The statement raises a possible competitive response but does not show that customers have switched suppliers.
NVIDIA reported record fiscal 2027 third-quarter revenue of $115.844 billion and GAAP net income of $65.287 billion, driven by Blackwell Ultra and Vera Rubin. Data Center revenue reached $107.61 billion. TechNews reported that the shipped Vera Rubin NVL72 platform contributed nearly $20 billion in incremental quarterly revenue, a figure not independently confirmed in the supplied material. A separate report said NVIDIA shares fell after the results and that the decline also affected neocloud companies CoreWeave and IREN, but provided no detailed explanation for the market reaction.
NVIDIA announced expanded commitments with supply-chain partners to align manufacturing capacity with demand for its accelerated-computing and AI infrastructure platforms. New agreements with Kinsus and Samsung Electro-Mechanics support additional substrate capacity and expanded production at existing factories, while work with Amkor and SPIL expands advanced-packaging capacity for current and next-generation platforms. NVIDIA also said Vera Rubin NVL72 systems and Vera CPUs began volume shipments in the third quarter, using TSMC’s 5nm and 3nm processes and CoWoS packaging.
Nvidia issued a sales forecast above Wall Street estimates in its first earnings report after former CEO Jensen Huang’s departure. Bloomberg and Bloomberg The Close described the outlook as a sign that AI-chip demand remained strong and that investor concerns about a larger impact on the AI industry had eased.
Nvidia interim CEO Colette Kress said the chip designer’s revenue would rise 75% year over year in the 12 months ending January 31, 2028, compared with the prior fiscal year. The forecast is five percentage points higher than the one Kress gave three months earlier.
The Information reports that JPMorgan and Citigroup lent money in 2025 to a Southeast Asian data center operator that did business with Megaspeed, which is at the center of a scandal involving Nvidia and former CEO Jensen Huang. The $2.8 billion financing arrangement raises questions about how thoroughly the banks vetted borrowers’ major customers, complex ownership structures and other potential red flags. The report does not identify the operator or specify how the financing was divided between the banks.
Nvidia reported fiscal Q3 revenue of $115.8 billion, up 103% from a year earlier and above the $109.1 billion estimate, according to Nvidia Newsroom. Data Center revenue reached $107.6 billion, up 110% year over year. Nvidia also forecast fiscal Q4 revenue above estimates, but the report did not specify the forecast amount or comparison basis.
According to an IT之家 report citing Nvidia’s financial results, fiscal 2027 third-quarter revenue was $115.844 billion, up 103% year over year, and attributable net income was $65.287 billion, up 105%. Revenue for the first nine months totaled $293.681 billion, while attributable net income reached $183.297 billion. The report said Nvidia’s board authorized an additional $150 billion in share repurchases and provided next-quarter revenue guidance of $130 billion. The article noted that its additional analysis was AI-generated and should not be treated as verified editorial analysis.
Nvidia told investors during its first earnings update since Jensen Huang’s departure that he is not coming back. The company also argued that it will continue to thrive without its longtime leader, according to Bloomberg Technology.
NVIDIA reported revenue of $115.8 billion for the third quarter of fiscal 2027, up 20% from the previous quarter and 103% year over year. Data Center revenue reached $107.6 billion, rising 21% sequentially and 110% annually. The company said the Vera Rubin NVL72 computing platform had shipped. Colette Kress, interim CEO and executive vice president and chief financial officer, said the product roadmap was on track.
At SC26, Nvidia introduced Nemotron for Science, the Vera CPU rack-scale platform, Rubin HPC, new CUDA-X libraries, quantum integration for Slurm, and photonics work with Lambda. Nemotron for Science combines Nemotron 3.5 Super VL with LitmusBench v0.2, nearly 50,000 chemistry-reasoning questions, and post-training recipes for domain-specific science agents. SiliconANGLE interprets the portfolio as infrastructure for agents that reason through scientific problems, invoke simulations, examine results, and choose the next step.
A live-coverage headline addresses Nvidia’s earnings results and the fallout involving Jensen Huang. The supplied report contains only the headline and provides no earnings figures, guidance, management commentary, financial period, or details about the nature of the Huang affair.
Nvidia is preparing to report earnings for the first time since Jensen Huang’s surprise departure, bringing its 70% growth outlook and the broader AI trade back into focus. Bloomberg Technology said compute demand remains strong despite leadership uncertainty, while investor concerns include growth and China.
Jensen Huang said he would be honored to return as Nvidia CEO, according to the original report’s headline. No further details were available in the supplied report.
What problem it solves
AI agents need to choose among models while balancing task success, cost, and latency for their workloads. Applications also need flexibility in where routing runs and who controls model clients, credentials, retries, and service operation.
Key features
Multiple routing algorithms — Routing can be configured with Auto, Task, Execution, or Composite strategies, including mode…
What problem it solves
Serving large language models requires efficient memory management, flexible model execution, and practical integration with existing models and APIs. vLLM addresses these needs with inference and serving capabilities spanning batching, caching, quantization, distributed execution, and multiple hardware platforms.
Key features
Efficient memory and serving — vLLM manages atte…
What problem it solves
AI agents need product-specific instructions to use NVIDIA software reliably across physical AI, robotics, simulation, CUDA-X libraries, RAG, and platform workflows. The catalog centralizes these skills while keeping them maintained in their respective product repositories and synchronized through an automated pipeline.
Key features
Agent-discoverable skill catalog — The NVI…
What problem it solves
AI agent skills need to be checked for quality, security, duplication, and their actual effect on agent behavior. Static checks alone cannot establish whether guidance improves task results, so evaluation must also support live comparison with and without a skill.
Key features
Three-tier evaluation pipeline — Independent tiers validate skill safety and structure, detect over…
What problem it solves
Running large language and vision-language models can require substantial setup and may depend on the available hardware. The project aims to make inference accessible across a wide range of local and cloud hardware with minimal setup.
Key features
Minimal-setup inference — The project provides installation options for common platforms, including scripts, Docker, pre-built b…
What problem it solves
Serving language models requires reading checkpoints, tokenizing requests, and running accelerated kernels while managing a local model server. TensorFold addresses these requirements with a native binary that can run on Apple Silicon and NVIDIA GPUs without requiring Python or MLX for serving.
Key features
Native Metal and CUDA inference — The Zig-based engine executes mode…
What problem it solves
Interactive multimodal and speech systems often require generation to span heterogeneous stages with different compute patterns, dependency structures, and resource needs. Serving these pipelines requires coordinated stage lifecycles, inter-stage transport, model integrations, and a unified API surface.
Key features
Coordinated multi-stage runtime — Models generation as coor…
What problem it solves
Data analysis workflows often require expressive query composition, flexible execution models, and interoperability with columnar data. Polars addresses these needs by allowing the DataFrame API and SQL to share the same optimizer and engine, with support for large workloads and custom extensions.
Key features
Unified DataFrame and SQL Queries — Queries can be written with e…
What problem it solves
Running frontier-scale open-weight models locally requires an inference platform that can make efficient use of the heterogeneous hardware available on a personal device. The project addresses this by treating GPUs, CPUs, host memory, and interconnects as a unified, elastic inference platform.
Key features
Edge-native MoE runtime — The runtime provides MoE serving with bandw…
What problem it solves
Autonomous AI agents often need to read files, install packages, call APIs, and use credentials, but unrestricted access can expose host data, secrets, and networks. OpenShell provides these capabilities within an isolated runtime where declared policies constrain agent access.
Key features
Kernel-level runtime enforcement — Agents run in isolated sandboxes where kernel cont…
What problem it solves
AI agent skills execute with implicit trust and often receive minimal vetting before installation. SkillSpector helps users assess whether a skill contains vulnerabilities, malicious patterns, or other security risks before installing it.
Key features
Multi-format scanning — Scan agent skills supplied as Git repositories, URLs, zip files, directories, or individual files.
So…
What problem it solves
Mosaics obscure portions of videos and images, making it difficult to view the affected content without changing the rest of the picture. Users need a restoration workflow that can identify selected mosaic regions and process them with GPU-accelerated video handling.
Key features
Targeted mosaic restoration — Detects mosaics in videos and restores them while leaving the rest…
What problem it solves
Deep learning models often need to be adapted for efficient deployment across different inference frameworks. Model Optimizer addresses this by providing a unified set of optimization techniques and export workflows for quantized checkpoints.
Key features
Post-training quantization — Compress models with quantization techniques available for Hugging Face, Megatron-Bridge, Di…
What problem it solves
Platform teams need to create and operate isolated Kubernetes clusters for customers on infrastructure they supply and control. They need a management plane that can provision GPU machines, manage cluster lifecycle operations, and establish node identity before issuing join credentials.
Key features
Independent workload clusters — Create workload clusters through the NVKE AP…
What problem it solves
Running AI models locally can require integrating multiple inference engines, model formats, hardware accelerators, and API surfaces. LocalAI addresses this with a small core, on-demand backends, and compatible APIs that work across supported hardware without requiring a GPU.
Key features
Composable inference backends — Backends are packaged separately and pulled only when a…
What problem it solves
LLM and Visual Gen inference requires specialized kernels, efficient runtime orchestration, and flexible deployment across NVIDIA GPU configurations. Developers also need a modular framework that supports customization while integrating with established inference tooling.
Key features
Optimized inference kernels — Custom kernels accelerate common inference operations, includ…
What problem it solves
Rust developers need host-side access to the CUDA runtime and tools for expressing GPU kernels while preserving Rust's ownership and borrowing rules. The platform provides separate programming models for workloads that require direct thread-level control and workloads naturally expressed through tensor operations.
Key features
Shared CUDA runtime — Host-side crates provide d…
What problem it solves
Agents can retrieve documents and query databases, but unreliable results undermine agentic analytics. Teams need a context layer that records where data lives, what its definitions mean, whether it is usable, how it should be queried, its provenance, and related concepts.
Key features
Enterprise ontology generation — Generate semantic ontologies from database metadata, infe…