Unverified50% confidenceFactExact time
Ollama's OLLAMA_MAX_LOADED_MODELS default value is 3 times the number of GPUs (3 for CPU inference)
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/7/2026
First Seen
Valid until: 12/6/2026
Sources
Related Entities
Related Claims
UnverifiedIn Ollama's CPU inference mode, generation speed drops from 30–80 tokens per second on GPU to 2–8 tokens per second.74% similarUnverifiedNPU 在执行 INT8/INT4 量化模型推理时，每瓦性能可达 GPU 的 3-5 倍72% similarUnverified该处理器集成40个RDNA 3.5计算单元的GPU核心和算力达50 TOPS的XDNA 2专用NPU67% similarUnverifiedOllama supports quantized lightweight models such as Phi-3 and Gemma 2B, which can be deployed on consumer-grade GPUs or even CPU-only edge servers.67% similarUnverifiedRyzen AI Max+ 395集成多达40个RDNA 3.5计算单元的GPU核心和算力达50 TOPS的XDNA 2专用NPU67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/870163API
curl https://kongchang.com/api/v1/knowledge/claims/870163MCP
get_claim(id=870163)