Abstract

The growing demand for artificial intelligence applications, including machine learning and deep learning, has led to increased reliance on high-performance processors…

Keywords

Liquid Cooling, AI Workloads, GPUs, TPUs, Immersion Cooling, Direct-to-Chip Cooling, Data Center Sustainability

Conclusion

Liquid cooling supports denser hardware deployments and extends hardware lifespan by offering significantly higher thermal transfer capacity and improved energy efficiency…

References

1. Choquette, J., et al. “NVIDIA A100 Tensor Core GPU: Performance and Innovation,” IEEE Micro, 2021.
2. Jouppi, N.P., et al. “TPU v4: An Optically Reconfigurable Supercomputer for Machine Learning,” ISCA, 2023.