repoGitHubTrust 82 · PrimaryPublished 1mo agoLive · yesterday
kekzl/imp
From-scratch C++23/CUDA inference engine for the NVIDIA RTX 5090 (sm_120a). The best single-GPU backend for agentic AI: tool calling, long-context loops, reasoning and concurrent sub-agents. Decode beats llama.cpp b9976 by 42-48% on dense GGUF (measured 2026-07-12), at-or-ahead of vLLM on NVFP4. 100% written by Claude Code.
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- PossiblePossibly related (embedding) · 57%NVIDIA Vera CPU Opens the Way for Agentic Scientific AI at Los Alamos National Laboratory →
- PossiblePossibly related (embedding) · 57%Nvidia’s AI Hardware Comes to Windows in RTX Spark PCs →
- PossiblePossibly related (embedding) · 56%Claude Meets Blackwell Ultra: Anthropic’s Models Now Run on NVIDIA GB300 in Azure →
- PossiblePossibly related (embedding) · 56%NVIDIA BioNeMo Agent Toolkit Brings Accelerated AI to Life Sciences Researchers in Claude Science →
- PossiblePossibly related (embedding) · 55%Build real agentic apps using CUGA: two dozen working examples on a lightweight harness →
- PossiblePossibly related (embedding) · 52%WattGPU: Predicting Inference Power and Latency on Unseen GPUs and LLMs →
- PossiblePossibly related (embedding) · 58%Top Cost-Effective Enterprise GPU Cloud Platforms for AI Workloads with H100–GB200, Elastic Scaling and Pay-as-You-Go Compute - Scott Coop →
- PossiblePossibly related (embedding) · 54%Anthropic’s Claude Available in Microsoft Corporation (MSFT) Foundry Powered by Nvidia GPUs - Insider Monkey →
Covers
newsNVIDIA Vera CPU Opens the Way for Agentic Scientific AI at Los Alamos National LaboratorynewsNvidia’s AI Hardware Comes to Windows in RTX Spark PCsnewsClaude Meets Blackwell Ultra: Anthropic’s Models Now Run on NVIDIA GB300 in AzurenewsNVIDIA BioNeMo Agent Toolkit Brings Accelerated AI to Life Sciences Researchers in Claude SciencenewsBuild real agentic apps using CUGA: two dozen working examples on a lightweight harness
Implements (incoming)
Covers (incoming)
newsTop Cost-Effective Enterprise GPU Cloud Platforms for AI Workloads with H100–GB200, Elastic Scaling and Pay-as-You-Go Compute - Scott CoopnewsAnthropic’s Claude Available in Microsoft Corporation (MSFT) Foundry Powered by Nvidia GPUs - Insider MonkeynewsAI Innovators Adopt NVIDIA Vera — Why Max Single-Threaded CPU at Scale MattersnewsUltra budget 20GB vram with 448GB/s for $100 bucks.newsNVIDIA Says AI Decoder Achieved up to 347 X Cut in Quantum Logical Error Rates - The Quantum InsidernewsNVIDIA Introduces New Jetson Thor Computers to Advance Mainstream Robotics and Edge AInewsWhy the first GPU financiers are turning to inference chips in a $400 million dealnewsNVIDIA DeepStream 9.1 is now open source: 13 new AI tools - en.softonic.comnewsT-Tech Open-Sources T-Search: High-Performance Agentic Retriever for Multi-Step Search That Runs on a Single GPU - quasa.ionewsNVIDIA Released DeepStream 9.1: Bringing Agentic AI to Vision AI With 13 Skills and Multi-View 3D Tracking - MarkTechPostnewsPSA: DO NOT use Intel consumer platforms for multi-GPU setupsnewsKog is going deeper to squeeze more inference out of GPUs
Related across the graph
newsNVIDIA Says AI Decoder Achieved up to 347 X Cut in Quantum Logical Error Rates - The Quantum InsidernewsAI Innovators Adopt NVIDIA Vera — Why Max Single-Threaded CPU at Scale MattersnewsNVIDIA Introduces New Jetson Thor Computers to Advance Mainstream Robotics and Edge AInewsUltra budget 20GB vram with 448GB/s for $100 bucks.newsBuild real agentic apps using CUGA: two dozen working examples on a lightweight harnessnewsPSA: DO NOT use Intel consumer platforms for multi-GPU setupsnewsClaude Meets Blackwell Ultra: Anthropic’s Models Now Run on NVIDIA GB300 in AzurenewsNVIDIA BioNeMo Agent Toolkit Brings Accelerated AI to Life Sciences Researchers in Claude SciencenewsAnthropic’s Claude Available in Microsoft Corporation (MSFT) Foundry Powered by Nvidia GPUs - Insider MonkeypaperWattGPU: Predicting Inference Power and Latency on Unseen GPUs and LLMsnewsNvidia’s AI Hardware Comes to Windows in RTX Spark PCsnewsNVIDIA Released DeepStream 9.1: Bringing Agentic AI to Vision AI With 13 Skills and Multi-View 3D Tracking - MarkTechPostnewsWhy the first GPU financiers are turning to inference chips in a $400 million dealnewsNVIDIA Vera CPU Opens the Way for Agentic Scientific AI at Los Alamos National LaboratorynewsNVIDIA DeepStream 9.1 is now open source: 13 new AI tools - en.softonic.comnewsT-Tech Open-Sources T-Search: High-Performance Agentic Retriever for Multi-Step Search That Runs on a Single GPU - quasa.ionewsKog is going deeper to squeeze more inference out of GPUsnewsTop Cost-Effective Enterprise GPU Cloud Platforms for AI Workloads with H100–GB200, Elastic Scaling and Pay-as-You-Go Compute - Scott Coop