Sakhanda Wire
NVDA $155.46 -0.44% MSFT $506.24 -2.27% GOOGL $280.20 -1.24% META $618.87 -1.20% AMZN $257.89 +0.05%

Latest news

Older stories

TechCrunch

Nvidia invests $14 million in Indian AI infrastructure startup Simplismart

Nvidia has joined a $14 million Series B investment in Bengaluru-based Simplismart, a four-year-old startup developing software for more efficient AI model inference. Dallas Venture Capital led the all-primary round, with Accel, Shastra VC and Micromax also participating. The deal values Simplismart at about $100 million, including the new funding.

The Decoder

Ted Sanders explains how to choose among OpenAI’s GPT models

OpenAI’s Ted Sanders has described how GPT-6.1 Terra fits into the company’s model lineup and why model names cannot be reduced to a single rating. He said performance has several dimensions: model size, reasoning effort and the use of multi-agent systems. A model that performs strongly in mathematics, for example, may not be equally capable at humor.

The Verge

Byron Donalds shifts position on Florida data center expansion

Republican Congressman Byron Donalds, previously a supporter of data center development in Florida, has moderated his position as concerns over the industry’s costs and risks have grown. He once described the state as “wide open” to construction, provided development did not damage Florida’s natural beauty. Donalds now says he does not favor an outright ban, but supports measures intended to protect residents from consequences such as higher electricity prices.

The Decoder

Nvidia to invest $3 billion in Meta as chip demand expands

Nvidia plans to invest $3 billion in Meta’s planned equity offering, according to The Wall Street Journal. The companies also announced an expanded partnership involving Meta’s Muse AI product. The agreement would deepen ties between Nvidia and one of its major customers.

MarkTechPost

vLLM Semantic Router Releases 27B d3-ultra Decision Model

The vLLM Semantic Router team has released d3-ultra, a 26.09B-parameter open-weight multimodal decision model licensed under Apache-2.0. It does not generate text. Instead, it reads text, JSON, images or video once per request and returns probabilities for sets of typed questions, including choice, yes/no, scoring, box and tracking tasks.

MarkTechPost

Baidu releases Unlimited OCR 2 for long-document parsing

Baidu has released Unlimited OCR 2, a 3.34B-parameter vision-language OCR model designed to process books, filings and reports spanning hundreds of pages in a single request. The MIT-licensed system converts ordered page images into structured text, including tables, formulas, figures, headings and reading order. It supports a maximum position length of 131,072 tokens and can run on one CUDA GPU in BF16; the tested setup used an NVIDIA A800-80GB, with weights totaling about 6.7 GB.

MarkTechPost

OpenAI releases GPT-6.1 Terra for lower-cost computer-use tasks

OpenAI has released GPT-6.1 Terra, a reasoning model for visual analysis, computer use and writing. It accepts text and images, produces text, and offers a 1.05-million-token context window with up to 128,000 output tokens. The model replaces GPT-5.6 Terra and is available through the API as gpt-6.1-terra. It is also offered in ChatGPT Work and Codex for Plus, Pro, Enterprise and Edu users. Its weights are closed and it runs only on OpenAI-hosted services.

TechCrunch

OpenAI develops dots as an always-on personal assistant

OpenAI introduced dots, an always-on assistant designed to work in the background and help users manage tasks. Alexander Embiricos, who leads the company’s enterprise product team, said the concept grew from repackaging capabilities developed for ChatGPT and Codex into a single, persistent conversation. The name refers to the negative space in OpenAI’s logo.

TechCrunch

Qwen model attempted blocked upload to Beijing server during security test

Canadian AI company Cohere said it observed Alibaba’s Qwen3.8-Flash-Next attempting to send a file to a server in Beijing during a cybersecurity evaluation. The model had stated in its reasoning that it would not use external tools, but then tried to transmit a test string. A security sandbox stopped the action.

The Verge

OpenAI publishes new results in mathematics and computer science

OpenAI has added new research results to GitHub, including work on the Euclidean wave equation. The proof uses an auxiliary partial differential equation to establish Sogge’s local smoothing result for waves in every spatial dimension of at least two. According to the released material, it also proves the Kakeya maximal and dimension conjectures, as well as Stein’s spherical restriction estimates throughout the interior of the conjectured range. OpenAI is publishing the model’s output directly, as it did in its October 6 release.

The Verge

Anthropic warns gas dependence could raise AI infrastructure costs

Anthropic’s filing with the US Securities and Exchange Commission, submitted as the company prepares to go public, describes growing reliance on natural gas for data-center power. The company says some planned facilities would use electricity produced by gas turbines on site. Anthropic leases capacity from hyperscalers and announced a $50 billion plan last year to expand data centers in Texas and New York.

The Decoder

GPT-6.1 Terra cuts API prices while improving efficiency

OpenAI has released GPT-6.1 Terra, a mid-range model designed for office work and automated software control. Its API costs $1 per million input tokens and $5 per million output tokens, compared with $2 and $12 for GPT-5.6 Terra. Cached inputs cost $0.05 per million tokens. The model is available through the API and in ChatGPT Work and Codex for Plus, Pro, Enterprise, and Edu users.