CUDA Expert

To lead the design and optimization of state-of-the-art GPU kernels powering advanced deep learning applications and large-scale workloads.

• You will drive cutting edge GPU kernel development and optimization across diverse, high-impact use cases.
• You’ll work at the intersection of hardware and algorithms – accelerating complex workloads and shaping the future of high-performance AI computation.
• Design, implement, and optimize GPU kernels for diverse deep learning workloads
• Collaborate with research and engineering teams to deliver state-of-the-art performance on modern GPU architectures
• Profile, debug, and fine-tune large-scale training and inference pipelines
• Stay ahead of the curve on GPU trends and emerging CUDA technologies

• Proven experience in GPU programming and parallel computing
• Hands-on experience in GPU kernel development for deep learning workloads (within the past 3 years)
• Strong understanding of modern GPU architectures – particularly Invidia Blackwell (B200) OR Hopper (H100)
• Familiarity with CUTLASS (Cuda framework)
• Proficiency with GPU profiling, debugging, and performance optimization tools
• Excellent analytical and problem-solving abilities

• Experience optimizing large-scale workloads
• Familiarity with Torch
• Familiarity with modern CUDA frameworks ( CuTe, CUB)
• Background in AI model training and inference optimization