typon

I used to think that Oxide's business model sucks in the face of the hyperscalers. When AWS/GCP/Azure "just works" and is generally reliable and cheap, why would I go through the trouble of buying my own computers? Well the past 5 years have been a constant decline towards more concentration of power, lack of care for their customers, and degradation in quality in general. AI is of course accelerating this decline - selling de-slopped products is now a huge competitive advantage. All the best luck to Oxide.

The LLMs "sweeping up" easy, or previously missed, results seems like a net negative. It's probably better for humans to struggle and come up with new tools than to just "clean up" low hanging fruit that doesn't add much value to the field.

* Software inference optimizations

* Heavy quantization

* Chips with hardcoded transformer architecture

* Much cheaper HBM

* Much sparser models - 1T total with ~1-10B active params e.g.

* Not to mention - 2 years of today's frontier models writing RTL and kernels at superhuman levels.

"we use a rough estimate of a total 9M GPU hours"

From CoreWeave, at current prices (~$2.46/hr spot to ~$6.16/hr on demand) would correspond to $22M–$55M.

The dataset is really where the cost is though - they used LVD-1689M - 1.6B images of curated web data from roughly 17B instagram images. This probably cost a huge amount of hours in human annotation, compute for algorithmic filtering, etc and not to mention probably a 20-50 person team working on this model.

You might want to change assumptions about how expensive these models are.

can you find curves like this for any other power source?

also batteries are getting exponentially cheap too

500 million users. Just because you haven't heard it doesn't mean an alternative doesn't exist.