jaeseok614
An open-source tool that calculates AI infrastructure configurations and estimates from available GPUs and goals.
SITE CAPTUREBased on information published on the service's website. Selection and review policy
WHAT IT DOES
What does it do?
It calculates which AI models can run and their expected speed based on VRAM and bandwidth. It distributes LLMs, embeddings, rerankers, and OCR/VLM workloads across available GPUs, showing allocation by GPU, recommended concurrent connections, expected throughput, and bottlenecks. Its results are calculated estimates that need validation before production use.
KEY FEATURES
Key features
- Calculate runnable AI models and expected speeds from GPU specifications
- Provide 3 GPU, CPU, RAM, storage, and network configurations and estimates
- Allocate LLM, embedding, reranker, and OCR/VLM workloads to GPUs
- Calculate recommended concurrency, expected throughput, and bottlenecks
- Show calculation grounds separately