nvidia - The LGTM (Page 4)

The LGTM

Home
Agentic Coding
Claude Code
Codex

Sign in Subscribe

nvidia

A collection of 74 posts

Local Coding Agents Are Becoming a Cluster-Sizing Problem

Local Coding Agents Are Becoming a Cluster-Sizing Problem

The least interesting phrase in the NVIDIA forum post is “22-agent dev team.” The internet already has plenty of prompt-based org charts pretending to be software companies. The interesting part is what happens underneath when those agents actually run in parallel: suddenly local coding assistants stop being a UX experiment

08 May 2026 5 min read

Claude Code’s Higher Limits Are Really a NVIDIA GPU Capacity Story

Claude Code’s Higher Limits Are Really a NVIDIA GPU Capacity Story

Claude Code did not get more generous because Anthropic discovered a new product-management philosophy. It got more generous because Anthropic found a very large pile of GPUs. That is the useful read on Anthropic’s new SpaceX compute deal, which gives the company access to more than 300 megawatts of

08 May 2026 5 min read

NCCL Inspector Turns Distributed Training Debugging Into a Dashboard Problem

NCCL Inspector Turns Distributed Training Debugging Into a Dashboard Problem

Distributed training failures have a special talent for making every layer look guilty. The model team sees throughput drop and blames the cluster. The cluster team sees green nodes and blames the dataloader. The network team asks for packet counters. Someone says “NCCL” in the tone usually reserved for old

08 May 2026 4 min read

FP8 Quantization Gets Practical for Multimodal Models on NVIDIA Hardware

FP8 Quantization Gets Practical for Multimodal Models on NVIDIA Hardware

Quantization has spent too long being marketed like a free lunch with a smaller datatype. NVIDIA’s latest Model Optimizer walkthrough is useful because it does the opposite. It shows FP8 post-training quantization as a workflow with calibration data, fake quantization, layer exceptions, benchmark evaluation, and an export path to

08 May 2026 4 min read

Rack-Scale AI Needs a Scheduler That Understands the Rack

Rack-Scale AI Needs a Scheduler That Understands the Rack

The quiet problem with rack-scale AI is that the rack stopped being furniture. NVIDIA’s GB200 NVL72 is sold as a single rack-scale AI system: 36 Grace CPUs and 72 Blackwell GPUs tied together by a 72-GPU NVLink domain with 130 TB/s of aggregate NVLink bandwidth. That makes for

08 May 2026 4 min read

NVIDIA’s IREN Deal Makes the AI Factory a Financing Strategy

NVIDIA’s IREN Deal Makes the AI Factory a Financing Strategy

The most important word in NVIDIA’s latest AI infrastructure announcement is not “gigawatts.” It is “right.” NVIDIA did not simply announce that IREN might build a lot of data center capacity around NVIDIA systems. It secured a five-year right to buy up to 30 million IREN shares at $70

07 May 2026 4 min read

NVIDIA’s New Ethernet Pitch Is Not ‘Faster Pipes.’ It Is Fewer Wasted GPUs.

NVIDIA’s New Ethernet Pitch Is Not ‘Faster Pipes.’ It Is Fewer Wasted GPUs.

NVIDIA’s latest Ethernet announcement is easy to misread as another throughput flex. It is not. Multipath Reliable Connection, or MRC, is really about avoiding the most expensive failure mode in modern AI infrastructure: synchronized training jobs waiting while a network fabric tries to remember how to behave. That sounds

07 May 2026 4 min read

NVIDIA’s Corning Deal Is the Supply-Chain Version of ‘Copper Does Not Scale’

NVIDIA’s Corning Deal Is the Supply-Chain Version of ‘Copper Does Not Scale’

NVIDIA’s Corning partnership is the supply-chain version of a technical admission: copper does not scale forever. The AI factory story has spent years focused on GPUs, memory bandwidth and software moats. Now the bottleneck is getting embarrassingly physical. At some point, the future of trillion-dollar AI infrastructure depends on

07 May 2026 5 min read

Why Agentic AI Sessions Break Conventional Inference — and What NVIDIA's Extreme Co-Design Stack Actually Fixes

Why Agentic AI Sessions Break Conventional Inference — and What NVIDIA's Extreme Co-Design Stack Actually Fixes

Here is a number that should concern anyone running or buying AI infrastructure today: a single live Claude Code session, tracked end-to-end over 33 minutes, made 283 inference requests — 58 from the main agent and 225 from sub-agents — and processed 3.5 million tokens before the first context compaction event.

05 May 2026 6 min read

The Car Is Now the AI Factory: NVIDIA's In-Vehicle Agent Stack Is Real and Running

The Car Is Now the AI Factory: NVIDIA's In-Vehicle Agent Stack Is Real and Running

Here's a sentence that would have sounded absurd five years ago but now describes a real product: you can buy a car this year that runs a full Nemotron-based agentic AI stack locally, routes complex queries to cloud inference when it makes sense, speaks to you through Magpie

05 May 2026 5 min read

Jensen Huang Says NVIDIA Has Zero Percent China Market Share — and Calls the Export Policy That Got It There 'Largely Backfired'

Jensen Huang Says NVIDIA Has Zero Percent China Market Share — and Calls the Export Policy That Got It There 'Largely Backfired'

Jensen Huang doesn't often admit to a zero. So when he stood in front of a camera and said NVIDIA now holds zero percent market share of AI accelerators in China — down from a dominant position just two years ago — it's worth taking seriously as a

05 May 2026 3 min read

NVIDIA's 90% Asian Supply Chain Exposure Is Now a First-Order Risk Factor, Not a Footnote

NVIDIA's 90% Asian Supply Chain Exposure Is Now a First-Order Risk Factor, Not a Footnote

When your cost structure looks more like a geopolitical risk map than a standard procurement spreadsheet, you've moved past the point where "supplier diversification" is a box to check. That's where NVIDIA finds itself today, with approximately 90% of its production costs flowing to

05 May 2026 3 min read

← Newer Posts Page 4 of 7 Older Posts →

The LGTM © 2026

Sign up

Powered by Ghost