Videos

Nvidia's reported talks to buy or invest in Reflection AI are part of a pattern of deal structures that mix...

Memory capacity and bandwidth, not GPU model numbers, decide what a card can run for local AI. The RTX 5090's 32 GB...

Loose Box's $5,999 AMD-based system claims 3.4x faster inference on Deepseek V4 Flash than Nvidia's DGX Spark, but...

The 64GB DGX Spark at $4,999 is a poor value compared to the Mac Mini M5 Pro, which offers faster memory and costs...

Nvidia's Open Agent Safety Platform is a layered answer to AI agents overriding their own safety instructions. Open...

An untouched one-bit Qwen 3.8 27B jumps from 29% to 61.5% on MMLU Pro and cuts time-to-first-token from 12.27 to...

NVIDIA is releasing a 64 GB DGX Spark at $4,999, halving the VRAM of the original 128 GB model while keeping the...

Switching matrix operations from Nvidia GPUs to Huawei's Ascend requires far more than a chip swap — the data...

Alibaba's Zhenru 900 beats Nvidia's H200 on paper in exactly two published specs, 216GB memory versus 141GB and a...

Nvidia's Nemotron 3 diarization model (roughly 100 million parameters, 4 GB GPU minimum) handles up to eight...

Docusign and NVIDIA built a purpose-built small VLM (Neotron Parse, ~850M parameters) that extracts complex tables...

The RTX 5090 carries a $3,400 premium over MSRP due to FOMO and export demand, making it the worst value for local...

Nvidia's open-source PAIR router allows agents to distribute inference across multiple local machines, each handling...

NVIDIA PAIR is not a GPU pooling tool that combines memory across machines. It is a request-level load balancer that...

ByteDance and Tsinghua's CUDA Agent beats PyTorch's compiler on 97% of 250 benchmark tasks, but the claim that it...

A GitHub repo with over 50,000 stars aggregates free API keys from 50 providers, giving Claude Code users 1.3...

OpenAI's custom inference chip, Jalapeno, designed with Broadcom in 13 months, shows preliminary benchmarks that...

Apple's new Mac Studio M5 Ultra (512GB, 1.2TB/s) is the first Apple product the speaker considers a genuine DGX...

NVIDIA paid $6 billion for a non-exclusive license to Poolside's model factory and 109 of its employees, not for the...

FreeToken's routing-aware expert caching cuts cache misses to 16% versus 62% for llama.cpp's static split,...

Nvidia's NOOA (Object-Oriented Agents) framework argues that most agent framework machinery—prompt templates, tool...

Nvidia's Nemotron 3.5 Lightning is 88% not a transformer — it's a hybrid of Mamba state-space layers and...

Nvidia dropped Nemotron 3.5 Lightning, a 30-billion-parameter mixture-of-experts model with only 3 billion active,...

NVIDIA quietly released Nemotron Lightning, a small open-weights MoE model (30B total, 3B active) built specifically...

NVIDIA's Switchboard is an open-source routing library that sits between your agent and its models, deciding...

NVIDIA's research shows that KV cache, previously thought model-specific, can be transferred between models using a...

Model routing is an emerging field where systems intelligently delegate tasks between large frontier models and...

AMD's MI355X GPU matches or beats Nvidia's B200 on paper but loses by a third in benchmarks due to software issues...

A debate is intensifying between proponents of open-source AI, led by Nvidia's Jensen Huang, and closed-source...

NVIDIA's Nemotron-3 models come in Nano, Super, and Ultra variants sized for consumer hardware, workstation/server...