Bitdeer AI, a division of Bitdeer Technologies Group (NASDAQ: BTDR) and a leading provider of AI Cloud services, has launched its state-of-the-art AI Training Platform. This new platform offers rapid and scalable AI/ML inference through serverless GPU infrastructure. With this release, Bitdeer AI becomes one of the first NVIDIA Cloud Service Providers (CSPs) in Asia to deliver both cloud services and an AI training platform.
The Bitdeer AI Training Platform enables users to build, train, and refine AI models efficiently using notebooks and organized project resources. Featuring pre-configured guides and customizable parameters, the platform simplifies AI model development and accessibility. It also facilitates collaborative model building across teams within the same organization, eliminating the need for server management and setting a new standard for efficiency and performance.
Cutting-Edge AI Infrastructure
The newly launched platform provides effortless access to advanced AI infrastructure, including NVIDIA DGX SuperPOD with H100 GPUs, DDN Storage, and InfiniBand Networks. It enhances the efficiency and scalability of AI/ML training by leveraging multi-GPU setups across multiple servers. By distributing tasks among several GPUs, Bitdeer AI’s services efficiently manage complex and large-scale training operations, making it an ideal solution for organizations looking to fast-track their AI projects.
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.
