1 min read
White Paper: GPU-Initiated, Liquid-Cooled, Ultra-High-Density Storage for Next-Gen AI
This paper introduces a paradigm shift in storage architecture designed to overcome the CPU-centric data path bottlenecks in modern AI workloads. By...
1 min read
Press
Updated on May 28, 2024
Nowadays, tremendous amounts of data processing, storage, and transmission are required for rising technologies and applications, such as cloud computing, artificial intelligence (AI), 5G mobile networks, and self-driving vehicles. To achieve high performance and efficient data processing, in addition to miniaturizing integrated circuits (IC) and increasing transistor density, semiconductor manufacturers use the three-dimensional integrated circuit(3DIC) technology that provides higher power density to boost computing performance. Yet, the 3DIC brings the challenge of high heat flux dissipation and causes some issues. For example, heat transfer problems will limit the essential demand in high processing speed; an abnormal temperature will increase failure probability and even reduce the service life of electronic components since the inefficient heat dissipation from electronic components. However, by deploying an excellent thermal solution can effectively solve these thermal issues.
This white paper investigates the thermal performance of different engineered designs of boiler plate surfaces and bases in immersion cooling. In the former sections, the engineered surface design (mesh, powder, and pin fin) and different bases (copper base and vapor chamber) are experimented with in low to high heat source thermal performance.
Wiwynn conducts several optimization solutions for improving the thermal performance of the boiler. The above experimental results are suitable for this test environment, and the results can serve as a reference for peers and customers. Wiwynn will continue innovating immersion cooling technology and provide advanced cooling solutions for the cloud industry.
Register to Download the whitepaper!
1 min read
This paper introduces a paradigm shift in storage architecture designed to overcome the CPU-centric data path bottlenecks in modern AI workloads. By...
1 min read
This paper explores a disaggregated key-value (KV) storage architecture designed to efficiently offload KV cache tensors for generative AI workloads.
1 min read
This paper explores an advanced framework designed to automate the extraction of important attributes from unstructured part datasheets. By...