Google, PyTorch 모델을 TPU에서 실행하는 TorchTPU 오픈소스 공개 ↗
Google이 Meta의 PyTorch 팀과 협력해 개발한 TPU용 PyTorch 백엔드 TorchTPU를 공개 . 기존 PyTorch 모델의 코드 변경을 최소화하면서 Google Cloud TPU에서 학습과 추론을 실행 GPU에서 쓰던 PyTorch의 실행 방식을 유지 하며, 연산을 하나씩 실행하는 즉시 실행 모...
공식 출처에서 수집한 소식을 발행일 순서로 확인하세요. 발췌는 원문의 일부이며 TUMN이 작성·검증한 해설 글과 구분됩니다. 전체 내용은 원문에서 읽을 수 있습니다.
2210개 · 25페이지 · 로그인 없이 열람
Google이 Meta의 PyTorch 팀과 협력해 개발한 TPU용 PyTorch 백엔드 TorchTPU를 공개 . 기존 PyTorch 모델의 코드 변경을 최소화하면서 Google Cloud TPU에서 학습과 추론을 실행 GPU에서 쓰던 PyTorch의 실행 방식을 유지 하며, 연산을 하나씩 실행하는 즉시 실행 모...
OpenAI says it's adding Orchestra's team and technology to work on Rosalind, its AI model for life sciences research, and to its related Rosalind Workbench tool, a harness that aims to help researchers plan experiments and analyze results.
SpaceX announced revenues of $14.6 billion in its Q3 earnings on Tuesday, which it says was a growth of 161 percent year-over-year "fueled by our Super Intelligence (SI) segment and accelerated growth in Connectivity." While its Space division posted a $7 million loss in income from operations, its …
Meta changes AI scraping policy but won’t add language respecting robots.txt.
It's probably time to audit your Gemini settings.
Google is bringing Gemini's Personal Intelligence feature to cars, allowing its AI assistant to draw on information from Gmail, Calendar, Maps, YouTube, and other Google services to deliver more personalized responses while driving.
OpenBSD 8.0 lands with Intel Arc support, ARM64 hibernation, and 64-vCPU guests
<p>My latest vibe coded game experiment is a Command & Conquer style RTS (Real Time Strategy) game called <a href="https://github.com/simonw/battledrones#readme">Battle Drones</a> (<a href="https://simonw.github.io/battledrones/">play it here</a>). Unlike <a href="https://simonwillison.net/2026/…
Comments
Microsoft is reportedly planning to launch an Xbox "next-generation handheld device" for the holiday season of 2027. The Xbox next-gen console, codename Project Helix, would then launch in the holidays of 2028, according to a new report from Fortune. The report doesn't make it clear whether the hand…
Google has joined Amazon in releasing a developer tool that makes it easy for PyTorch users to move their apps off Nvidia GPUs.
Mistral이 음성 인식과 합성 모델군 Voxtral 3와 음성 에이전트 SDK를 공개 . 전화나 웹에서 대화하며 정보를 검색하고, 청구서를 확인하거나 예약을 처리하는 서비스 구축 가능 한국어를 포함한 50개 언어의 음성 전사 와 화자 구분 지원. 전문 용어와 참고 문맥을 ...
Meta says it paused a Muse feature that lets you share conversations with the AI assistant after researchers found the URLs publicly available on the web, as reported by Ars Technica. In a report published this week, researchers at the security firm Claroc say they found more than 6,800 Muse links t…
The Smartsheet Model Context Protocol (MCP) Server lets AI agents read, update, and act on live data. This post shows two ways to connect it on AWS: Amazon Quick for business teams and Kiro for developers, so your agents can work with Smartsheet without switching tools.
The Smartsheet Model Context Protocol (MCP) Server lets AI agents read, update, and act on live work management data. This post shows two ways to connect it on AWS: Amazon Quick for business teams and Kiro for developers, so your agents can work with Smartsheet without switching tools.
The former Google CEO owns about 1% stake in the leading AI lab, Bloomberg reported.
Tuesday Labs, a New York City-based robotics lab and design studio, has opened reservations for Voxel, its new bi-pedal robot designed to learn and take over routine tasks throughout your home. Similar to other robots like Weave Robotics' Issac 1 that are designed for household chores beyond mopping…
Learn how we fine-tuned GPT-OSS-20B for multilingual reasoning on Amazon SageMaker HyperPod, and how to decide between supervised fine-tuning (SFT) and reinforcement learning with GRPO. SFT took reasoning-language accuracy from 22% to 100%, and GRPO added measurable gains where headroom existed.
Comments
Meta argues nobody “decided” to pirate books to train AI.
Octave built a secure, multi-tenant AI foundation on Amazon Bedrock, using Cohere Embed v4 for embeddings and Amazon Bedrock AgentCore for isolated agent execution, to power a Complaint Triage Assistant that searches 2 billion regulated documents without crossing tenant boundaries.
Meta's AI companion Muse has surpassed 10 million U.S. downloads in its first 40 days, joining an exclusive group of apps that includes Threads. New data also shows Muse is retaining users at higher rates than its AI rivals.
Comments
Canadian organizations can now access OpenAI GPT-5.6, GPT-6, and GPT-6.1 models on Amazon Bedrock from the Canada (Central) and Canada West (Calgary) Regions using cross-Region inference. This post covers the models and inference profiles, how to set up permissions, and how to invoke them with the R…
GPT-4.1로 만든 허위 통계 페이지 5만 개 를 게시한 실험 사이트가 7개월 동안 ChatGPT 유입 세션 1만 건 이상과 330개 이상의 도메인으로부터 링크를 얻음 Google 검색 노출이 급감한 뒤에도 ChatGPT의 조회와 인용은 지속됨 . 최근 30일간 ChatGPT가 답변 중 페이지...
The post Microsoft opens a hands-on STEM and AI hub for students at its Grange Castle data center campus appeared first on Source .
Hardcoded tool wiring breaks down as an agent's catalog grows from a handful of tools to hundreds. This post walks through a runtime Discover, Instantiate, Execute loop built on AWS Agent Registry, where an orchestrator discovers MCP servers and A2A agents through semantic search at runtime instead …
AMD has just revealed that it's bringing its latest upscaling technology, FSR 4.1, to the Xbox Series S and X consoles, as well as laptops, 6000-series GPUs, and handhelds like the Xbox Ally X devices. FSR 4.1 will start appearing for Radeon RX 6000-series GPUs today, as well as Ryzen AI Max 300 and…
It’s the moment everyone is waiting for — the ChatGPT moment for robotics, Amulya Vishwanath explained to TechCrunch’s Equity.