## Blog posts

### June 11, 2025

**Adaptive Engine Certified for NVIDIA GBX B200: Adding Support for Blackwell Architecture**

Announced at [NVIDIA GTC Paris](https://www.nvidia.com/en-eu/gtc/?ncid=partn-861353-1), Adaptive ML is proud to be officially certified for the NVIDIA GBX™ B200, and join NVIDIA’s ecosystem of enterprise AI partners as [an NVIDIA-Certified System](https://www.nvidia.com/en-us/solutions/ai-factories/validated-design/).

Equipped with eight NVIDIA Blackwell GPUs interconnected with NVLink™, [NVIDIA DGX B200](https://www.nvidia.com/en-us/data-center/dgx-b200/) delivers 3x the training performance and 15x the inference performance of the NVIDIA DGX H100, making it an ideal choice for enterprises looking to efficiently fine-tune, evaluate, and serve specialized models with [Adaptive Engine](/content/engine/index.html).

The NVIDIA DGX B200 is a unified AI factory, melding cutting-edge hardware with the complete NVIDIA AI software stack, including [NVIDIA Base Command](https://www.nvidia.com/en-us/data-center/base-command/?ncid=no-ncid) and [NVIDIA AI Enterprise](https://www.nvidia.com/en-us/data-center/products/ai-enterprise/?ncid=no-ncid) software, now including Adaptive Engine’s reinforcement fine-tuning, synthetic data generation, and custom evaluation capabilities.

Armed with the NVIDIA Blackwell architecture, DGX B200 can handle diverse workloads, including reinforcement learning, recommender systems, guardrail models, custom chatbots, and more.

Watch [Jensen Huang’s announcement keynote](https://www.youtube.com/watch?v=X9cHONwKkn4) to learn more, or [get started with Adaptive Engine](/content/book-a-demo/index.html).