Search “NVIDIA”

esults on AA-AgentPerf for Meta's new open-weight model Muse Spark Open, an 816B-parameter, 42B-active MoE that scores 39 on the Artificial Analysis Intelligence Index. Tests show NVIDIA B300 and GB300 each reach about 150M tokens/s per megawatt under throughput-first configurations, with GB300 at about 108M and B300 at about 92M at 100 tok/s output speed; AMD MI355X reaches 70M tokens/s per megawatt at 65 tok/s. 46 of the model's 62 layers use 2,048-token sliding window attention, reducing KV c

Matched fieldsSummary、Body

that small neoclouds find it difficult to build a complete GPU deployment team on their own, and Dell's managed services support is key to keeping clusters healthy and profitable after they go live. Alpha Compute provides sovereign and confidential computing with Dell and Intel hardware; its 72-node St. Louis cluster uses Dell's Nvidia Blackwell systems, InfiniBand networking and direct liquid cooling, and it has signed a $31.9 million non-recourse loan agreement secured by Nvidia B300 hardware.

Matched fieldsSummary、Body

um3, training Qwen3-32B on 16 Trainium2 chips; eager mode reached 11,099 tokens/s and 23.89% MFU, while compiled mode reached 14,156 tokens/s and 30.47% MFU. The author considers the Neuron Explorer profiler superior to Nvidia's and AMD's solutions, but first-time compilation takes too long: the first-step compilation for the gpt-oss training recipe took 27 minutes 17 seconds. FlexAttention tests passed 89 of 96 cases, with all 7 failures due to compilation timeouts rather than numerical errors.

Matched fieldsSummary、Body

heise conducts an efficiency test on the Nvidia RTX Pro 6000 Blackwell, measuring its performance loss at three power levels of 450 watts, 300 watts and 156 watts; the card has 96 GByte VRAM and targets AI and workstation scenarios. The test uses 450 watts as a reference corresponding to the previous-generation GeForce RTX 4090, 300 watts to the long-term limit of earlier flagship cards, and 156 watts to mid-to-low-end cards such as the GeForce RTX 5060 or Radeon RX 9060 XT. The full test data i

Matched fieldsSummary、Body

Matched fieldsSummary、Body

Matched fieldsSummary、Body

tealth, closing a $2.2 million pre-seed funding round led by Ground State Ventures, with participation from Vanagon Ventures, Plug and Play, and Freiraum Ventures. Its platform Rhyme provides an abstraction layer that lets developers with basic Python skills write and debug code for quantum computers; the compiler translates the code for quantum hardware and supports multiple quantum processors. RhymeQ is an NVIDIA Inception member, and its technology is compatible with NVIDIA's CUDA-Q platform.

Matched fieldsSummary、Body

Matched fieldsSummary、Body

silicon architecture, advanced materials, and connectivity technologies are long-term advantages that are hard to replicate. Israel has deep accumulated strength in chip design, photonics, and high-speed communications, and entrepreneurs can target the architectural bottlenecks encountered by giants such as Nvidia, Apple, Intel, and Google. Commercialization success depends on yield, reliability, and system integration, and early validation with industrial buyers and design partners is crucial.

Matched fieldsSummary、Body

MiniMax open-sources M3.1 Flash, a 428B-parameter MoE with 23B activated parameters, gaining day-one support from Together, SGLang, vLLM and NVIDIA NeMo. OpenAI's Decisions API is officially available, powered by GPT-6 Luna, with classification and decision speed improved by up to 10x, at an input price of $0.10 per million tokens. Anthropic's Claude models are officially available on Google Cloud, and starting next week, Claude sessions on Pro, Max and Team plans will run in auto mode by defaul

Matched fieldsSummary、Body

MSI has released the EdgeMesa N AI+ mini PC, with a volume of 1.3 liters, equipped with the NVIDIA RTX Spark N1X superchip and up to 128GB of LPDDR5X unified memory, with running 120B-parameter large models locally as its primary design goal. The chip integrates a 20-core Grace CPU and a Blackwell GPU with 6144 CUDA cores; the unified memory pool lets model weights and KV cache share the same physical memory. Referring to measured results from the DGX Spark on the same platform, generation speed

Matched fieldsSummary、Body

At Groundbreak 2026, construction management software company Procore Technologies announced a deepened collaboration with NVIDIA to bring digital twin and simulation capabilities based on NVIDIA Omniverse into complex construction projects; the feature is currently in closed beta. Users can switch between Procore's Model Manager and Realism Mode to simulate in a high-fidelity 3D environment how real-world conditions such as extreme weather affect construction; the collaboration is mainly aimed

Matched fieldsSummary、Body

of the M3.1 Flash model, which processed over 100 trillion tokens within 72 hours of launch. Compared with the previous generation M3, M3.1 Flash has comprehensively improved capabilities, with usage cost reduced by 70% and online inference speed increased by nearly 3 times. MiniMax stated it will work with chip manufacturers such as Huawei Ascend, NVIDIA, and AMD, as well as inference frameworks and communities such as vLLM, SGLang, and Hugging Face, to advance day-one open-source integration.

Matched fieldsSummary、Body

Matched fieldsSummary、Body

Mifeng Technology, together with nearly 20 institutions including Hugging Face, NVIDIA, and Tsinghua University, jointly launched the embodied evaluation platform RoboColiseum, and announced this round's results on October 20: Claude Opus 5.5 ranked first overall among the four flagship models, followed closely by GPT-6 Astra. GPT-6 Astra scored 99.35 on instruction following, while Opus 5.5 performed better on spatial understanding and general manipulation and consumed the fewest tokens. The pl

Matched fieldsSummary、Body

Matched fieldsSummary、Body

The facility has a total floor area of approximately 36,878 square meters, with 9 floors above ground and 2 below, and is scheduled for completion in 2029. All server rooms in the facility will support liquid cooling, and it will use a high-load structural design to accommodate Nvidia's next-generation Vera Rubin GPUs, while also introducing a free-cooling system to reduce cooling energy consumption. The project is part of its long-term strategy to secure 1GW of AI data center capacity by 2031.

Matched fieldsSummary、Body

s within a single connector, and the company claims it can arrange 1,000 fibers within 800mm2. The company has laid out a three-step roadmap of NPO optical engines, CPO, and optical circuit switches. Its first NPO module is targeted for launch in 2028 or 2029, and CPO products are expected to be released by the end of this decade. If Nvidia's Vera Rubin NVL576 is estimated at one pair of fibers per SerDes, it would require approximately 20,736 pairs of fiber connections to connect 8 NVL72 racks.

Matched fieldsSummary、Body

Cohere announced that the confidential computing feature of Model Vault has entered public beta, providing hardware-level isolation for hosted model inference based on the NVIDIA H100 GPU confidential computing capability on Microsoft Azure. This feature extends the scope of encryption from data at rest and in transit to prompts, documents, responses, and model execution during inference; after end-to-end encryption on the application workload side, decryption can only occur in a confidential en

Matched fieldsSummary、Body

Matched fieldsSummary、Body