此语言仍有 24 条内容或界面任务待翻译;未完成部分显示中文原文。

Insights & News

AI、算力、具身智能机器人与智能硬件的产业动态,以及选型与采购指南。

今日更新0 篇Updates This Week23 篇Published147 篇订阅周报
AI Chips & StorageIndustry Updates

Biren Technology Says It Has Completed Inference Adaptation of MiniMax H3-Image on Biren-bring GPU

According to Beijing Business Today, Biren Technology (HK6082) posted on 11月11日 that MiniMax released MiniMax H3-Image and opened its weights on the evening of 11月10日. Building on the SGLang inference framework and the BIRENSUPA software stack, Biren completed model adaptation on its Biren-bring series general-purpose GPU and offers developers a deployment solution. The report did not disclose performance or deployment-scale data.

AI Chips & StorageIndustry Updates

Moore Threads Announces MDC 2026 Developer Conference in Beijing on 29 12, Next-Generation GPU Progress in Focus

Moore Threads (688795) disclosed on the evening of 11 11 that the 2026 MUSA Developer Conference (MDC2026) will be held in Beijing from 29 to 30 12, with more than 10 new products, technologies, and solutions expected to be announced. Progress on the next-generation GPU product has become a key focus. Founder Zhang Jianzhong will deliver a keynote speech. According to earlier reports, the S6000 AI computing card and the "Lushan" series based on the next-generation "Huagang" architecture are expected to launch by year-end, and the conference will also reveal further progress on the C128/C256 supernode.

AI Chips & StorageIndustry Updates

Enflame Technology's Three-Day Price Deviation Exceeds 30%, Triggering Abnormal Fluctuation Announcement

Enflame Technology (688801) announced that its shares recorded a cumulative deviation of closing price gains of 30% or more over three consecutive trading days from 2026-11-9 to 11, constituting an abnormal fluctuation under SSE rules. After self-inspection and consulting the actual controller, the company confirmed no undisclosed material matters, while disclosing revenue of RMB 23.32 billion and net profit attributable to the parent company of RMB -7.54 billion for the first three quarters of 2026 — still unprofitable — and noting that tradable shares account for only 4.1595% of total share capital, with short-term share price volatility risk.

AI Chips & StorageIndustry Updates

PhiGent Robotics Showcases PhiGo Entry Parking-and-Driving Integrated Solution, with Cumulative Design Wins Exceeding 270 Units

According to Gasgoo, on 2026 11, 6, PhiGent Robotics presented its PhiGo Entry parking-and-driving integrated solution at the 2026 Evolution NavInfo Innovation Conference held in Shanghai. The solution targets vehicles in the RMB-1,000-class segment and overseas market demand, and has now launched the world's first large-scale mass-production deliveries, with cumulative design wins exceeding 270 units.

2026-11-11 · 本站更新

AI Chips & StorageIndustry Updates

NVIDIA Publishes Vera CPU Comparison Data: Leads in Both Agents per Watt and Speed vs x86

TMTPost reports that on 2026 11, 5, NVIDIA published Vera CPU comparison data on its official blog: under identical conditions in a 100-megawatt data center, Vera supports 1.6x more Agents per watt than the fastest x86 CPU, and delivers 2x faster Agent speed than the most power-efficient x86 CPU. Vera carries 88 in-house Olympus cores and 176 hardware threads, designed for Agent workloads; early users such as Perplexity, DeepInfra, and Daytona have published production-grade test numbers.

钛媒体 · 约 7 分钟

2026-10-27 · 本站更新

AI Chips & StorageIndustry Updates

Meta生产环境启用Google TPU,PyTorch原生TorchTPU方案落地

Meta、Google与RadixArk合作推出TorchTPU,使SGLang、vLLM等PyTorch推理引擎无需重写代码即可在TPU上运行;Meta确认PySGLang-TorchTPU已进入生产部署,减少对Nvidia生态的依赖。文中还给出Ironwood TPU与GB200的Kimi K3推理对比数据,以及Amazon开源TorchNeuron等相关动向。

同一事件另有 1 篇报道

飞翔科技频道(转载Toms Hardware) · 约 3 分钟

AI Chips & StorageIndustry Updates

Google TPUs Supplied to External Customers for the First Time, Now Running in Meta's Production Environment

Google's in-house Tensor Processing Units have been supplied to an external customer for the first time and are now handling production AI workloads in Meta's data centers. The key to deployment is inference engine adaptation: SGLang has released a PyTorch-native version for TPUs, and vLLM has also published a corresponding open-source version, allowing most models to migrate to TPUs without code rewriting.

IT时代网(腾讯新闻发布页) · 约 4 分钟

AI Chips & StorageIndustry Updates

Yuanjie Semiconductor Announcement: Projected Revenue of 6.5 to 7亿元 for Q3 of 2026, First Batch of Phase II Equipment to Enter Plant Soon

Shaanxi Yuanjie Semiconductor Technology Co., Ltd. issued an announcement on its recent production and operations, projecting Q3 2026 revenue of 65,000.00万元 to 70,000.00万元, up 264.56% to 292.60% year over year; and net profit attributable to the parent of 40,300.00万元 to 45,300.00万元. The company said AI computing power demand is driving demand for high-speed optical modules and upstream laser chips, CPO/NPO-related products are advancing steadily, and the Phase II facility is expected to begin forming capacity from the end of 2026 to early 2027.

东方财富网(转载源杰科技公告) · 约 5 分钟

2026-10-24 · 本站更新

AI Chips & StorageIndustry Updates

NVIDIA Partners with Six Supply Chain Partners to Expand Substrate, Packaging, Testing, and Optical Engine Capacity

NVIDIA announced a joint capacity expansion with six supply chain partners—IBIDEN, Samsung Electro-Mechanics, Unimicron, Amkor, KYEC, and SPIL—across four areas: packaging substrates, advanced packaging, chip testing, and CPO optical engines. Substrate capacity is targeted to reach 2.5 times current levels within three years, interposer wafer capacity to increase by more than 300% by the end of 2028, chip testing capacity to double by 2027 (backed by a USD 11 million/billion prepayment), and optical engine capacity to expand to 15 times current levels by mid-2027. The expansion is primarily located in Taiwan (China), Japan, South Korea, and the United States, with capacity coming online in phases starting 2027.

同一事件另有 2 篇报道

大广网 · 约 6 分钟

2026-10-15 · 本站更新

AI Chips & StorageIndustry Updates

微软发布新一代自研AI芯片Maia 300:2027年推出,推理能效提升40%

微软宣布下一代自研AI加速器Azure Maia 300将于2027年推出,并从2027年起将Maia系列部署扩展到数吉瓦规模,作为主要推理系统。芯片配备288GB内存、18TB/s带宽,每兆瓦Token生成量较领先GPU系统提高40%。微软暂不对外出租Maia算力,优先满足自身AI服务需求。

澎湃新闻(腾讯新闻转载页面) · 约 4 分钟

2026-10-04 · 本站更新

硅基之家示例内容 · 约 1 分钟

Product Selection Assistant

让需求变成可修改的筛选条件

依据已录入参数生成条件,结果可查看和修改。帧率与性能需核对实测,不能仅凭算力数值保证。