1 min read
White Paper: CAE-Assisted Cable Routing Methodology for High-Density Server Platforms
As AI server systems continue to scale rapidly in power and data throughput, cable density within server platforms has increased significantly....
1 min read
Wiwynn
Updated on July 11, 2025
As thermal design power (TDP) for modern processors such as CPUs, GPUs, and TPUs exceeds 1 kW, traditional air cooling methods are proving insufficient. Liquid cooling, particularly using cold plates, has emerged as a superior solution. This paper from Wiwynn introduces an advanced cold plate design that integrates in-package liquid cooling with jet impingement techniques to minimize junction-to-liquid thermal resistance, eliminating multiple layers of thermal interface material (TIM) and improving heat transfer efficiency.
The proposed integrated cold plate was evaluated through numerical simulations and experimental setups. Simulation results show that these integrated cold plates outperform traditional designs, reducing thermal resistance by 8%. Additional fluidic channels within the chip package, along with direct liquid cooling, further enhance performance by up to 25.7%. The jet impingement cold plate design significantly improves thermal resistance by 20–48% compared to traditional cold plates. When combined with the integrated cold plate, the impingement design reduces thermal resistance by up to 53%.
These advancements demonstrate that integrated impingement cold plates offer significant improvements in thermal management for high TDP processors, making them a promising technology for data centers and AI applications.
Register to Download the whitepaper!
1 min read
As AI server systems continue to scale rapidly in power and data throughput, cable density within server platforms has increased significantly....
1 min read
As server power density continues to increase in modern AI and high-performance computing applications, direct liquid cooling (DLC) has become a...
This whitepaper introduces SPDM attestation between a BMC and an External Root of Trust (EROT) on AI servers. It outlines SPDM prerequisites and the...