老黄花7亿美元收购的LeptonAI，创始人贾扬清被曝仅一年就离职了。 就在刚刚，SemiAnalysis爆出猛料：LeptonAI创始人、CEO贾扬清已从英伟达离职。 距离老黄斥7亿巨资完成收购，刚好一年零两个月。 如今，不仅核心产品DGX Lepton没打出水花，当初承诺2026年前开源Lepton核心软件平台，半个字都没兑现。 从Caffe到LeptonAI，开源二十年 贾扬清大家都很熟悉了 ...

Google’s going to the movies, as it invested $75 million in the hot indie studio A24, along with a pledge to provide AI to ...

Founded by the mind behind the Swift programming language, Modular’s 'write once, run anywhere' stack looks to accelerate ...

The TDN72 rack houses 288 Napier chips, delivers 608 petaflops of FP8 compute, and claims to match the throughput of nine NVIDIA Rubin plus Groq LPX racks.

如果你平时调大模型 API，大概率遇到过这几种情况：某个平台今天额度用完了，代码直接报错想试新模型，又要去另一个网站申请 Key，重复配置 base ...

Abstract: ZENO is an AI-powered voice assistant engineered as a standalone, interactive system capable of delivering intelligent, human-like conversations on embedded platforms. Developed using a ...

AI 快用不起了。大模型机构密集发出“涨价信号”。 在国内，智谱 4 月 8 日随 GLM-5.1 新模型发布同时宣布涨价 10% ——这是它 2026 年内第三次涨价，CEO 张鹏的解释很直白：“公司存在算力约束和瓶颈。”一周后，大洋彼岸的 Anthropic 出手： Claude 企业版原本 $200 ...

When Nvidia first showed off its Compute Unified Device Architecture (CUDA) parallel computing platform in 2006, it was a multibillion-dollar bet that failed to turn a profit for a decade. Today, it ...

NVIDIA Corporation CEO Jensen Huang has been deliberately emphasizing the agentic AI inflection in his recent commentary, which likely sets the tone for upcoming GTC 2026 revelations. A Groq-based LPU ...

openbench provides standardized, reproducible benchmarking for LLMs across 30+ evaluation suites (and growing) spanning knowledge, math, reasoning, coding, science, reading comprehension, health, long ...