Table of Contents |
---|
Containers
Container Purpose | Option 1 (GPU) | Option 2 (CPU) |
---|---|---|
Gateway (Linux) | 16 GB 4 vCPUs | 32 16 GB 8 4 vCPUs |
Embedder (Linux) | 16 GB 16GB GPU None. Embedding happens on Gateway.16 GB GPU 2 vCPUs | 16 GB 4 vCPUs |
LLM (Linux) | 16 GB 24 GB GPU 2 vCPUs | To be determined. |
Dashboardt3a.large/Ingestor (Windows Server- Not a container) | 8 GB 4 vCPUs | t3a.large |
Info |
---|
|
AI model servers
The main hardware cost of the BusinessGPT deployment are the AI servers responsible for answering questions. These utilize GPUs.
...