Bitdeer AI unveils advanced AI Training Platform

Bitdeer AI, part of Bitdeer Technologies Group, a leading Artificial Intelligence (AI) Cloud service provider, has announced the launch of its advanced AI Training Platform, designed to provide fast and scalable AI/ML inference with serverless GPU infrastructure. With the newest AI Training Platform, Bitdeer AI becomes one of the first NVIDIA Cloud Service Providers (CSP) in Asia to offer both cloud service and an AI training platform.

The Bitdeer AI Training Platform empowers everyone to build, train, and fine-tune AI models at scale through notebooks and organized resources on a project basis. Based on the pre-configured guides and customizable parameters, the innovative platform simplifies the process of developing and refining AI models, making them accessible to a wider audience. It further allows different teams within the same organization to collaboratively build and develop AI models without the need to manage their own servers, setting a new benchmark in efficiency and performance.

High-Performance AI Infrastructure

The newly announced platform offers seamless access to high-performance AI infrastructure and resources of NVIDIA DGX SuperPOD with H100 GPUs, DDN Storage, and InfiniBand Networks. It also improves the efficiency and scalability of AI/ML training processes by utilizing multi-GPUs across various servers simultaneously. By distributing the workload across several GPUs, Bitdeer AI’s services can handle extensive and sophisticated training tasks, making it the optimal choice for organizations aiming to accelerate their AI initiatives.

Addressing Key Business Challenges

• Optimizing Development Costs: With Bitdeer AI, businesses can optimize costs through a pay-as-you-go model, only being charged when notebooks are in service mode. This approach ensures that organizations only pay for the resources they use, making AI development more cost-effective.
• Simplifying Complex GPU Infrastructure Setups: The serverless infrastructure provides a comprehensive integrated development environment for ML, including pre-built algorithms and support for popular frameworks like TensorFlow and PyTorch. This significantly reduces the complexity and time required to develop and train ML models, streamlining the AI development process.
• Ensuring Reproducibility and Environment Consistency: Bitdeer AI ensures consistency and reproducibility in the build environment, crucial for managing ML model deployment. This consistency prevents unexpected errors when restarting CI/CD jobs or migrating from one platform to another, avoiding costly build errors in long-running ML jobs.

Bitdeer AI collaborated with a software engineering team from the SMU School of Computing and Information Systems to test, verify, and fine-tune the platform, ensuring its robustness and effectiveness. Looking ahead, Bitdeer AI plans to collaborate with NVIDIA to enhance the AI Training Platform by integrating with the NVIDIA AI Enterprise (NVAIE) cloud services such as NIM. This collaboration will enable businesses to customize, test, and scale AI agents efficiently, further solidifying Bitdeer AI’s commitment to providing top-tier AI solutions.