NVIDIA reported revenue of $96.22 billion for the second quarter of fiscal 2027, more than doubling from a year earlier and beating the $92.17 billion consensus compiled by LSEG. Adjusted earnings per share came in at $2.22 for the three months ended July 26, above the $2.10 expected by the market, while data center revenue rose to $89 billion, also ahead of forecasts. For the third quarter, NVIDIA projected revenue of $108 billion, plus or minus 2%, topping analysts’ average estimate of $104.19 billion. Its adjusted gross margin outlook was 74%, plus or minus 0.5 percentage point, below the market expectation of 74.77%. The company also said it expects fiscal 2028 revenue to grow about 70% and announced an expanded partnership with Amazon Web Services. CFO Colette Kress said NVIDIA and AWS plan to deploy an additional 2 million NVIDIA GPUs across Amazon’s global infrastructure during 2027 and 2028. NVIDIA said its AI chip revenue opportunity before 2027 could exceed $1 trillion. It added that its third-quarter guidance does not include any data center chip sales to China.

Nvidia shares recovered in after-hours trading on Wednesday after the company released its fiscal second-quarter results. The stock had closed the regular session down 1.59% at $209.66, then rose to about $217 after hours, roughly 3.5% above the close. The move followed a report showing quarterly revenue more than doubled to $96.2 billion, topping the $92.27 billion analyst estimate cited by the Associated Press. Data Center revenue climbed 117% to $89 billion, adjusted earnings came in at $2.22 per share against a $2.09 estimate, and gross margin held at 75%. The company also disclosed $366 billion in multiyear artificial intelligence infrastructure commitments, including $279 billion tied to supply and capacity, along with phased guarantees of up to $108.5 billion that it said reflect potential exposure rather than current losses or payments. Nvidia said it expects third-quarter revenue of $108 billion, plus or minus 2%, and forecast gross margin of 74%, plus or minus 0.5 percentage point. The earnings release also noted that shipments of Hopper data-center products to China accounted for less than 1% of Data Center revenue during the quarter.

Zhipu has launched GLM-5.3-Flash, a native multimodal large model, according to the Z.ai website. The company said the model has 320 billion total parameters and 18 billion active parameters, and that it performs clearly better than GLM-5.2 on coding and agent benchmarks. Zhipu also said the model comes close to Claude Opus 4.8 while cutting pricing to about one-tenth of that level. The company said GLM-5.3-Flash uses a hybrid sparse and linear attention architecture, Manifold-Constrained Hyper-Connections, and 30 trillion tokens of multimodal training data. On that basis, Zhipu said long-context inference costs were reduced to about one-third of GLM-5.3. In Artificial Analysis Intelligence Index v4.1.1, the model posted a score of 57 at roughly $0.045 per task. Zhipu added that during anonymous testing under the name ox-alpha on OpenCode and OpenRouter, the model became the most popular one within a week, and that it has already reached inference efficiency close to NVIDIA GPUs on large domestic AI chip clusters.

Bitdeer said in its July 2026 production and operations update that it mined 1,190 BTC during the month, up 322% from a year earlier. The company reported about 76.7 EH/s of self-mining hash rate, 5.0 EH/s of other proprietary hash rate, and 18.7 EH/s tied to joint mining through third-party data center operations, bringing total proprietary hash rate to 81.7 EH/s. As of the end of July, it held 257 BTC, excluding customer deposits. On the AI side, Bitdeer disclosed a 16-year hosting lease valued at $4.7 billion with a Volta subsidiary for an AI/HPC data center at the Tydal site in Norway. If a one-time eight-year extension option is exercised, the total contract value could reach $8 billion. The company said 121 IT MW at the site will be fully equipped with NVIDIA GPUs for a leading AI lab, with a PUE of about 1.1 and 100% renewable energy use. It also said about $1.3 billion in letters of credit support is expected to be arranged by affiliates of two international financial institutions, subject to customary conditions.

RUM Group, formerly Rumble Inc., has signed a six-year GPU services agreement with a U.S. cloud customer worth about $13.7 billion, according to Techub News, citing CryptoBriefing. The deal includes a warrant provision that gives the customer the right to buy more than 50.8 million Class A shares at $0.01 per share. Exercise terms are tied to actual service purchases and any potential expansion agreements. The contract will be carried out in three tranches of roughly $4.6 billion each, with final delivery dates subject to customer approval. The GPU services covered by the agreement will support the company’s Quake AI division. Techub also said RUM Group acquired about 85.2% of Northern Data AG and 22,000 NVIDIA GPUs in June this year. Based on the figures cited, the new contract is about 50 times larger than the company’s earlier $270 million agreement with Together AI.

Researchers from the University of California, Berkeley and the University of Texas at Austin have introduced FreeToken, an edge-native mixture-of-experts inference engine designed to turn personal computers into a unified elastic inference platform. According to the release cited by Techub, the system can run the 753B-parameter GLM-5.2 model on a single workstation GPU, a 284B model on a gaming desktop, and a 35B model at interactive speed on a laptop GPU with 8GB of VRAM. FreeToken has been open-sourced under the Apache-2.0 license on GitHub and published on PyPI. The team also provides one-click desktop applications for Windows and Linux. Its command-line interface supports Linux x86_64 systems and NVIDIA GPUs, and the ft serve command can expose an API endpoint on port 1919 that is compatible with OpenAI and Anthropic. The project is aimed at individual developers, startups, and engineering teams at small and medium-sized businesses, with a focus on privacy-sensitive use cases such as healthcare, legal work, defense, finance, and intellectual-property-heavy R&D. Example applications include local coding agents, private code review, offline contract analysis, and synthetic data generation.

Alpha Compute reported that cash revenue from its gaming business and the Alpha-01 NVIDIA GPU cluster reached $1.57 million in July. The company said that figure was about 35 times higher than its total revenue in April 2026. The update was carried by ChainCatcher under the technology category. No further breakdown of the revenue mix or additional operating details were disclosed in the brief. The report focused on two business lines: gaming operations and the Alpha-01 NVIDIA GPU cluster, both of which contributed to the July cash revenue total.

SemiAnalysis is framing SpaceX’s latest valuation case around delivery speed rather than raw GPU counts, arguing that the scarce asset in AI infrastructure is not only chips but the ability to bring large-scale compute online within months. In the analysis cited by WhiteLine and translated by Wu Blockchain, the firm says SpaceX can create value by preparing power capacity and data center space first, then buying newer-generation NVIDIA GPUs closer to energization. That setup lets the company avoid locking in older chips too early and direct near-term available capacity to the highest-paying customers. The report points to two contracts to illustrate the pricing logic. Google is said to pay about $920 million per month starting in October 2026 for roughly 110,000 NVIDIA GPUs and related compute resources, while Anthropic previously agreed to buy SpaceX compute for $1.25 billion a month. The article says customers are paying for immediate access because their own facilities are not yet online even as training and inference demand has already arrived. SemiAnalysis also outlines a more aggressive scenario in which SpaceX could approach 10 GW of compute capacity by the end of 2027 and generate about $300 billion in annual recurring revenue. Whether that holds depends on delivery speed, NVIDIA’s supply of newer GPUs, and contract renewals once customer-owned data centers begin operating.