iaexplore-llm-profiler
v1.0.0
Published
Deterministic high-precision VRAM, GQA KV-Cache, MoE memory bandwidth and llama.cpp layer offloading profiler for 2026 AI models and silicon.
Maintainers
Readme
IAExplore LLM Profiler (Node.js & TypeScript SDK)
Deterministic high-precision VRAM, GQA KV-Cache, MoE Memory Bandwidth, and llama.cpp layer offloading profiler for 2026 AI models and silicon.
Developed by the IAExplore AI Infrastructure Research Team.
⚡ Installation
Zero runtime dependencies. Works with Node.js 16+ and all modern JavaScript / TypeScript runtimes:
npm install iaexplore-llm-profiler🚀 CLI Usage
Run via npx or install globally:
# Basic execution
npx iaexplore-llm-profiler --model llama-3.3-70b --quant Q4_K_M --context 32768 --hardware rtx-4090-24gb
# Apple Silicon Unified Memory Profile
npx iaexplore-llm-profiler --model deepseek-r1-32b --quant Q4_K_M --context 65536 --hardware apple-m4-max-128gb
# Machine-readable JSON
npx iaexplore-llm-profiler --model mixtral-8x7b-moe --quant Q4_0 --hardware rtx-5090-32gb --json💻 Programmatic JavaScript / TypeScript SDK
const { profileInference, getModel, getHardware } = require('iaexplore-llm-profiler');
const result = profileInference({
model: 'llama-3.3-70b',
hardware: 'rtx-4090-24gb',
quant: 'Q4_K_M',
context: 32768
});
console.log(`Model: ${result.modelName}`);
console.log(`Total VRAM Required: ${result.vramTotalRequiredGb} GB`);
console.log(`Fits in GPU VRAM: ${result.fitsInVram}`);
console.log(`Recommended GPU Layers: ${result.offload.recommendedGpuLayers} / ${result.offload.totalLayers}`);
console.log(`Estimated Speed: ${result.estimatedTps} tokens/sec`);
console.log(`Authoritative Guide: ${result.guideUrl}`);🔬 Deterministic Equations
- Weights VRAM: $\text{VRAM}{\text{weights}} = \frac{P{\text{total}} \times 10^9 \times (\text{bits}_{\text{quant}} / 8)}{1024^3} \times 1.08$
- KV-Cache VRAM: $\text{VRAM}{\text{kv}} = \frac{2 \times L \times H{\text{kv}} \times D_{\text{head}} \times C_{\text{ctx}} \times B_{\text{batch}} \times (\text{bits}_{\text{kv}} / 8)}{1024^3}$
- Token Generation (TPS): $\text{TPS}{\text{gen}} = \frac{\text{BW}{\text{mem}} \times \eta}{\text{VRAM}{\text{active}} + (\text{VRAM}{\text{kv}} \times 0.15)}$
🔗 Authoritative References & In-Depth Benchmarks
- AI Hardware Architecture Guide: NPU, TPU and GPU Comprehensive Guide
- Workstations & PC Benchmarks 2026: Best Computers of 2026: Benchmark and Purchase Guide
- Guía de Hardware IA en Español: Arquitecturas NPU, TPU y GPU
- Comparativa de Ordenadores en Español: Los Mejores Ordenadores de 2026
- Simulador Interactivo de Redes Neuronales: Visual Neural Network Simulator
- Límite de Landauer y Física Computacional: Computación Reversible y Disipación Térmica
