this post was submitted on 01 Dec 2024
46 points (88.3% liked)

Futurology

1840 readers
358 users here now

founded 1 year ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
[–] kippinitreal 8 points 1 week ago (3 children)

Genuine question: how energy intensive is it to run a model compared to training it? I always thought once a model is trained it's (comparatively) trivial to query?

[–] [email protected] 7 points 1 week ago (1 children)

A 100-word email generated by an AI chatbot using GPT-4 requires 0.14 kilowatt-hours (kWh) of electricity, equal to powering 14 LED light bulbs for 1 hour.

Source: https://www.washingtonpost.com/technology/2024/09/18/energy-ai-use-electricity-water-data-centers/

[–] [email protected] 0 points 1 week ago

How much energy does it take for the PC to be on and the user to type out that email manually?

I assume we will get to a point where energy required starts to reduce as the computing power increases with moores law. However, it's awful for the environment in the mean time.

I don't doub that rather than reducing energy, instead they will use more complex models requiring more power for these tasks for the foreseeable future. However eventually it will be diminishing returns on power and efficiency will be more profitable.

[–] [email protected] 6 points 1 week ago (2 children)

For the small ones, with GPUs a couple hundred watts when generating. For the large ones, somewhere between 10 to 100 times that.

With specialty hardware maybe 10x less.

[–] pennomi 3 points 1 week ago (2 children)

A lot of the smaller LLMs don’t require GPU at all - they run just fine on a normal consumer CPU.

[–] [email protected] 3 points 1 week ago (1 children)

Wouldn't running on a CPU (while possible) make it less energy efficient, though?

[–] pennomi 3 points 1 week ago

It depends. A lot of LLMs are memory-constrained. If you’re constantly thrashing the GPU memory it can be both slower and less efficient.

[–] [email protected] 1 points 1 week ago

yeah but 10x slower, at speeds that just don't work for many use cases. When you compare energy consumption per token, there isn't much difference.

[–] kippinitreal 2 points 1 week ago

Good god. Thanks for the info.

[–] [email protected] 2 points 1 week ago

Still requires thirsty datacenters that use megawatts of power to keep them online and fast for thousands of concurrent users