Research Kernel Engineer

About Bitdeer

Bitdeer is a pioneering global technology firm focused on Bitcoin mining and AI cloud services. It delivers end-to-end Bitcoin mining solutions encompassing ASIC chip design, mining rig manufacturing, and the comprehensive computational value chain including equipment procurement, logistics, datacenter design and operation, as well as network and facility management. The company operates internationally with a diversified 3 GW energy portfolio and has deployed datacenters for Bitcoin mining and High-Performance Computing across multiple countries including the USA, Bhutan, Norway, Canada, Malaysia, and Ethiopia.

About Bitdeer AI Lab

Bitdeer AI Lab is an advanced research division within Bitdeer, dedicated to pushing AI frontiers with a long-term vision to make AI inference affordable and efficient. The Lab’s work involves innovating on both hardware and software infrastructure, aiming to optimize the cost and performance of AI model serving through infrastructure investments and novel technology development.

Key Responsibilities

• Accelerate research by creating novel efficient implementations of newly developed AI inference methods such as new quantization techniques, sparsity schemes, attention variants, and speculative decoding strategies that currently lack optimized counterparts.
• Develop and optimize CUDA and Triton kernels tailored for cutting-edge research methods to achieve tangible speed improvements on actual hardware.
• Conduct performance analysis throughout the AI serving pipeline using techniques like roofline modeling to identify bottlenecks and optimize kernel performance specific to Bitdeer's models and traffic patterns.
• Convert slow reference implementations of published methods into practical, production-ready solutions.
• Establish robust performance measurement frameworks to support ongoing optimization efforts.
• Collaborate closely with the AI Cloud platform team, aligning kernel development with the broader serving infrastructure.

Qualifications and Skills

• Possess a Bachelor’s, Master’s, or PhD degree in Computer Science, Electrical Engineering, or a related discipline with relevant hands-on experience in GPU programming, HPC, or ML systems.
• Proficient in CUDA and/or Triton, with demonstrable experience writing and optimizing kernels; strong programming skills in Python and C++.
• Solid understanding of GPU architecture and memory systems including occupancy, memory coalescing, tensor cores, and warp-level primitives, supported by familiarity with profiling tools like Nsight Compute or Nsight Systems.
• Practical experience optimizing critical inference components such as attention mechanisms, GEMM operations, normalization, sampling algorithms, or key-value cache management.
• Able to implement novel methods that lack existing reference implementations, working effectively from academic papers, experimental notebooks, or conceptual drawings.
• Capable of designing precise performance metrics to guide optimization and provide transparent interpretation of results.
• Experience with inference engines like vLLM, SGLang, or TensorRT-LLM and developing custom kernels or extensions for them is highly desirable.
• Exposure to compiler frameworks or intermediate representation tools (MLIR, TVM, TorchInductor), or distributed serving technologies such as tensor or pipeline parallelism is a strong plus.
• Contributions to high-tier systems research publications or significant open-source projects in GPU computing or inference are welcomed.
• Strong passion for cutting-edge AI infrastructure, exceptional ownership, and disciplined engineering practices.

Work Culture and Benefits

• An inclusive, authentic workplace that embraces diverse ideas and backgrounds.
• A collaborative environment with open office spaces and an energetic startup atmosphere.
• An opportunity to engage with top industry pioneers and contribute significantly to the digital asset space.
• Participation in pioneering projects, system process development, and hands-on impact creation.
• Supportive environment fostering autonomy, rapid growth, continuous learning, and personal accountability.
• Competitive welfare benefits along with training and mentorship programs for growth and development.