Nvidia Senior VP Gilad Shainer said Co-Packaged Optics (CPO) has entered mass production to meet the massive data throughput and speeds required by Nvidia’s Vera Rubin platform, which is at least 3x faster than GB300, media report, noting the bandwidth performance of scale-up (intra-rack/chip interconnection) will be 10x that of scale out (inter-rack data center networking). Switches developed by Nvidia and partners have already been sent to key clients and deployed at Nvidia. Shainer is head of R&D for Nvidia InfiniBand and Ethernet high-speed transmission technologies. The announcement is seen as Nvidia’s answer to industry researchers that have recently said broad deployment of CPO might be delayed until 2028-2029 due to heat challenges and complex ASIC integration. $NVIDIA(NVDA.US) $Taiwan Semiconductor(TSM.US) $Lumentum(LITE.US) $Coherent Corp.(COHR.US)
Source: Dan Nystedt
The copyright of this article belongs to the original author/organization.
The views expressed herein are solely those of the author and do not reflect the stance of the platform. The content is intended for investment reference purposes only and shall not be considered as investment advice. Please contact us if you have any questions or suggestions regarding the content services provided by the platform.