Falcon-180B (Technology Innovation Institute) and GPT-3.5 (davinci-002) (OpenAI) are both frontier AI models. Falcon-180B was published in September 2023 and GPT-3.5 (davinci-002) in March 2022.
Falcon-180B was trained on 3.8 x 10^24 FLOP, about 1.5x the compute of GPT-3.5 (davinci-002) at 2.6 x 10^24 FLOP. Training compute is the closest available proxy for how much was invested in a model, though it says nothing on its own about how well that compute was spent.
These two models share no benchmark on which both have been scored, so no direct performance comparison is possible here. The specification table below is a comparison of inputs, not of results.
| Field | Falcon-180B | GPT-3.5 (davinci-002) |
|---|---|---|
| Organization | Technology Innovation Institute | OpenAI |
| Published | Sep 6, 2023 | Mar 15, 2022 |
| Training compute | 3.8 x 10^24 FLOP | 2.6 x 10^24 FLOP |
| Parameters | 180B | — |
| Dataset size | 3.5T | — |
| Training hardware | NVIDIA A100 SXM4 40 GB | NVIDIA A100 SXM4 40 GB |
| Chips used | 4,096 | — |
| Training time | 4.3K h | — |
| Training cost (2023 USD) | $11M | $5M |
| Training power draw | 3.3 MW | — |
| Accessibility | Open weights (restricted use) | API access |
| Open weights | Yes | No |
| Country | United Arab Emirates | United States |
