Unverified50% confidenceEventExact time
When using stageload's release(module) to move parameters and buffers to the meta device and running Qwen-Image-2.1 in a stage-by-stage 'load at start, release at end' manner (the 16.3 GB text encoder and the 13.3 GB transformer loaded only one at a time), the peak on the same Mac dropped to 19.0 GB, and swap did not grow in either run.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/13/2026
First Seen
Valid until: 1/11/2027
Sources
Related Entities
Related Claims
UnverifiedWhen the author ran Qwen-Image-2.1 at 1024×1024 resolution on the same Mac, the process footprint never exceeded 18.5 GB, but swap grew by 3.5 GB during the denoising stage and by 9.6 GB during the VAE decoding stage, and finally the memory guard actively terminated the run before writing out the image.73% similarUnverifiedThe article uses a 125-billion-parameter Qwen model with a file size of 105GB; this task could only be run on the Mac Studio due to the model's size, as the 32GB Mac mini could not accommodate it.69% similarUnverifiedTesting with a 4-bit quantized Qwen2.5 7B model on an M2 Mac mini with 16 GB unified memory, throughput dropped from 16.8 tokens/s without batching to 8.7 tokens/s with batching.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1015603API
curl https://kongchang.com/api/v1/knowledge/claims/1015603MCP
get_claim(id=1015603)