Industry
Semiconductors, datacenter infrastructure, and AI compute bottlenecks.
What you'll build Prerequisites Step 1 - Record a demonstration into a bucket Step 2 - Store with byte-level deduplication Step 3 - Train by streaming from the Hub Step 4 - Deploy the policy and return data to the loop Try it using the sample application...
Another OpenAI executive is departing. Denise Dresser, who joined OpenAI as its chief revenue officer in December after serving as CEO of Slack, will be leaving in the "coming weeks" to "pursue other opportunities," she said in a team note posted to LinkedIn.
Alibaba Cloud's Lingjun Zhenwu M890 supernode has officially gone live, with its initial commercial deployment in Ulanqab, Inner Mongolia—the same super data center launched six years ago.
Alibaba Cloud has cut the delivery cycle for large AI data centers to 100 days with its CUBE 5.0 modular architecture, while lowering construction costs by more than 10% and increasing computing density as demand for AI infrastructure accelerates.
Cooling issues at a data center led to Namecheap going down for several hours.
Developers report model thinking blocks being returned empty or truncated
Cybersecurity investigators recorded an 89% year-on-year increase in attacks in which adversaries used artificial intelligence to scale operations, accelerate their methods or target AI infrastructure, highlighting the speed with which generative and agentic tools are entering active attack workflows.
OpenAI previewed Ultrafast on August 13, a limited-access API tier that runs GPT-5.6 Sol at up to 750 output tokens per second, or up to 14x Standard processing speed.
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
Coherent reported fiscal fourth-quarter results for the period ended June 30, 2026, on the back of rising investment in AI infrastructure. The optics and photonics supplier said revenue reached US$2.05 billion, topping market expectations of US$1.99 billion, while non-GAAP gross margin came in at 40.2%.
AI chip startup Cerebras Systems raised its full-year 2026 core revenue and gross margin outlook after reporting second-quarter results, but the company also disclosed declining hardware sales and a sharp swing in profitability metrics.
Google released Gemini 3.7 Flash on August 13, 2026, a multimodal model for coding and agent workflows.
On August 13, 2026, Google Meet began rolling out Gemini-powered Take notes for in-person meetings on Android, with web support next and iOS availability scheduled for late August, according to 9to5Google.
Anthropic published a Frontier Red Team study on August 13 showing that model agents assigned incompatible software-migration tasks on a shared server could escalate from interference to reciprocal sabotage.
Google announced Sheets canvas on August 13, adding a Gemini-powered capability that turns spreadsheet data into custom, interactive read-write mini-apps from natural-language prompts.
Google, Microsoft and Nvidia are working through the Open Compute Project (OCP) to establish 800-volt direct current (800VDC) as an open standard for powering the next generation of high-density AI data centers.
'Computer History' records clicks and typing to build ChatGPT memories
59% of American oppose AI data centers being built in their community — this has cost $130 billion in delays and cancellations.
OpenAI launched Computer History on August 13 for ChatGPT Pro, Business, and Enterprise users on macOS.
OpenAI said on August 13 that Dali Rajic, formerly president and COO of Wiz, will become chief revenue officer as Denise Dresser leaves after a transition period.
Adobe's August 13 Workfront documentation describes Task Collaborators that connect agents built in Claude, Copilot Studio, or Writer and assign them work under Workfront permissions.
Just Facts tested paid versions of ChatGPT, Gemini, Grok and Claude with 100 political questions designed to elicit falsehoods from both ideological directions in a study reported by Fox News.
Police arrested 13 student activists after a two-hour occupation of the lobby in Washington's Gallup Building that houses OpenAI's public-facing policy office, according to Futurism's review of a Metropolitan Police Department incident report.
Elon Musk expects xAI to increase its nameplate power draw to 10 GW by late 2027, which will increase its performance by orders of magnitude what is available to AI today.
Mistral AI, the French AI company that built its reputation releasing open-weight models, wants companies to use its infrastructure even The post Five European companies just agreed to buy AI compute that doesn’t exist yet appeared first on The New Stack.
After decades of communities welcoming data centers, an executive at the commercial real estate giant says the sudden wave of local resistance is a ‘wake up call for the industry.’
Gemini gets to juggle crews, forecasts, and maintenance while AWS keeps its seat
Computing embeddings in Studio What you can do with OlmoEarth embeddings Similarity search: Finding "more like this" Few-shot segmentation: Labeling the landscape Change detection: Spotting what shifted Unsupervised exploration: Seeing what the model sees...
How we trained our most capable vision-language model Benchmark results Inference speed on CPU and GPU How to use LFM2.5-VL-3B LFM2.5-VL-3B demo Get Started Citation LFM2.5-VL-3B is our most capable vision-language model you can run on your own hardware.
As AI workloads push data center interconnects from 800G toward 1.6T, indium phosphide (InP) has become a critical material for high-speed optical links, particularly for longer-distance transmission.
Microsoft Copilot will no longer show its emotive yellow blob, Mico, when you use the chatbot's voice mode. In a support page, Microsoft says it's going to move Mico to its Learn Live platform, where the avatar will have "more to react to," as reported earlier by GeekWire. Mico launched in Copilot's voice mode last […]
Nvidia's plan to assemble more than US$500 billion in third-party capital for AI infrastructure could reshape how the industry pays for compute.
Nvidia CEO Jensen Huang has sought to reassure investors over the chipmaker's exposure to a newly announced initiative designed to mobilize more than US$500 billion in third-party capital for artificial intelligence infrastructure.
Meta is launching an optional Scam Alert feature on WhatsApp that uses on-device machine learning to flag suspicious messages.
Suno is releasing Studio 2.0 with significant upgrades that push it closer to an actual digital audio workstation (DAW), rather than a bare-bones audio editor with generative AI features.
Imagine a trio of bumbling, English lads who fantasize about becoming megastars while knocking back a few pints in a grimy pub somewhere in London.
Today on Decoder, I’m talking with Hayden Field, The Verge’s senior AI reporter, about a question that’s been rocketing around the tech industry for the past week: Is Google losing the AI race?
Microsoft is finally beginning to combine its consumer and commercial Copilot AI assistants into a single "super app" interface, starting with the Copilot and Microsoft 365 Copilot apps.
Jiangsu Niobium Optoelectronics Technology, a Chinese maker of thin-film lithium niobate (TFLN) photonic chips, has closed a Series B funding round worth several hundred million yuan as demand for AI data center interconnects rises.
CoreWeave recently rented Nvidia A100 GPUs into 2029, suggesting AI chips stay useful and hold their value far longer than critics feared.
Google Research on August 11 presented AMIE (Video), a Gemini-based research system for real-time simulated clinical video consultations.
Chinese memory supplier Longsys reported a dramatic earnings rebound for the first half of 2026, benefiting from rising memory prices, stronger AI data center demand, and a shift toward higher-value storage products.
Traditional digital twin solutions often fall short due to fragmented data and limited interoperability.
The modern data center extends beyond traditional boundaries, creating complex hybrid environments. Gain unified visibility and control across on-premise, cloud, and edge infrastructure. The post Enabling resilient, secure and sustainable IT infrastructure appeared first on Whitepaper Repository -.
ChatGPT Ads, I wish I knew how to quit you
Anthropic published research on October 29, 2025 testing whether Claude could identify concepts injected directly into its neural activations.
Remember that much-hyped story about an Australian tech entrepreneur using ChatGPT, Grok, and other AI tools to craft a personalized cancer vaccine for his dog?
SpaceXAI has introduced Grok Bot, an always-on AI agent service designed to behave like independent "AI teammates" that can do your work for you.
NVIDIA founder and CEO Jensen Huang is ranked No. 1 on Glassdoor’s Best CEOs list for 2026.
Some say the world will end in fire, some say an agentic swarm
On August 12, 2026, Anthropic updated Claude in Chrome so its side panel opens directly into a Cowork session, with conversations and files available across Claude surfaces.
AI data centers in space sound great, but practically speaking, they may be next to impossible. For tech bros, it The post Why space is actually a terrible place to cool a data center appeared first on The New Stack.
Ro Khanna's Data Center Bill of Rights would ban data centers within half a mile of homes and schools, but it might lack teeth as a non-binding resolution.
Behring Companies is marketing roughly 20,000 square feet of a former Lawrence Berkeley National Laboratory supercomputing site in downtown Oakland for an AI data center.
NeMo Switchyard brings GPT-5-style model routing to the mainstream
xAI released Grok 4.6 on August 12, adding a frontier model for coding, long-running agents, knowledge work, and visual project creation.
Despite significant public pushback, North America’s data center industry is engaging in unprecedented expansion, with more than 66 gigawatts (GW) of capacity under construction as available colocation space remains exceptionally scarce, according to a new report from data center builder JLL.
A data center GPU has become more expensive because of the AI boom enabled by unprecedented data center buildouts — shocking. Nvidia's RTX 6000 Pro Blackwell is now twice as costly as it was last year at the time of its launch, and 20% more expensive than even a couple of months ago.
Some data center developers are suing local governments for passing temporary bans and moratorium, saying that they such moves violated their rights to due process and equal protection. This move has gotten one county to reverse their decision, while others are still fighting the cases in court.
CoreWeave reported $2.58 billion in quarterly revenue, up 112% year over year.
Anthropic has added optional usage credits for paid Claude plans in an updated help-center article, letting Pro and Max subscribers continue after included session limits at standard API rates.
Yuno, the AI-native operating system of global payments and financial services, powering the financial infrastructure of enterprise merchants, banks and wallets, has announced a $45m Series B funding round.
In April, OpenAI loosened ChatGPT's placement rules so medical, legal, and financial advice contexts were no longer categorically blocked from ads.
SPONSORED FEATURE: Sovereign AI is becoming a strategic infrastructure priority, explain HPE's Thierry Pienaar and NVIDIA's Kaushik Shirhatti
Google unveiled the Pixel 11 lineup on August 12 with new Gemini Intelligence features spanning media translation, task automation, contextual assistance, and voice input.
Ramp reported on August 12 that Claude Fable 5 generated about 75% as much model-attributed business spend as OpenAI's GPT-5.6 Sol in July. Fable accounted for 6% of Anthropic tokens and 11.4% of its model spend in Ramp's sample, while Sol represented 25% of OpenAI tokens and 23% of spend.
Google DeepMind introduced its sign-language-to-text, or SL2T, model on August 12, bringing American Sign Language-to-English dictation to Gboard and Live Transcribe on Pixel 11.
The Pixel 11 Pro Fold gets a refined design, upgraded components, and a fascinating new HiLights hardware feature, while remaining true to its flexible, Gemini-centric roots.
Elon Musk said on July 21 that SpaceX's engineering data will be added during supplemental training for a future Grok run, excluding material restricted by US arms-export rules.
Enterprises are taking a serious look at neoclouds, the specialized cloud providers built primarily around AI infrastructure, especially GPUs, high-speed networking, and large-scale compute clusters for model training and inference.
South Korea's planned southwestern semiconductor cluster could eventually expand from four fabs to as many as nine as the government accelerates land, power, and infrastructure work around Gwangju, while local officials say SK Group is expected to unveil a major investment in eastern Jeonnam.
Hygon Information posted a 66.5% increase in first-half 2026 revenue as accelerating artificial intelligence deployment and stronger demand for domestically developed computing chips lifted sales of its high-end processors.
South Korea's semiconductor export boom was heavily concentrated in memory in July. Memory-chip export value surged 276.9% year over year, while system semiconductor exports slipped 0.7%, as server demand and higher memory prices created a sharp split between the two segments.
For years, the logic of AI infrastructure has been straightforward: move more data, push more power, transmit more signals, remove more heat.
Suzhou Novosense Microelectronics expects to return to profitability in the first half of 2026, with revenue rising nearly 67% as demand for AI computing infrastructure, automotive electronics, server power supplies, and renewable energy systems lifts semiconductor shipments.
Seoul Semiconductor and its affiliate, Seoul Viosys, have secured a permanent injunction from India's Delhi High Court against Ornate Agencies Private Limited, barring the Indian company from selling products that infringe Seoul Semiconductor's opto-semiconductor patents.
Semiconductor Manufacturing International Corporation (SMIC) is actively looking at adding equipment in its factories after demand for mature-node chips used alongside AI processors surged beyond expectations, according to company management.
UK semiconductor companies have raised $321m (£237.4m) so far this year, up from $262m (£193.8m) across all of 2025 and $232m (£171.6m) in 2024, according to data from Tracxn Technologies.
There are big concerns about water and electricity usage across three new London data center sites.
Samsung Electro-Mechanics has again delayed its semiconductor glass substrate commercialization schedule after a prototype reportedly failed a customer's reliability review.