No abstract

A novel optical transceiver with 4 transmitter plus 4 receiver channels designed for coupling to multicore multimode fiber has been fabricated and characterized. The transceiver is based on the holey Optochip concept where 4-channel VCSEL and photodiode arrays are flip-chip attached to a single-chip SiGe IC using AuSn solder. Optical vias (holes) are fabricated into the SiGe IC to enable optical access to the conventional topside emitting 850-nm optoelectronic arrays. The optoelectronic arrays are arranged in a quad-VCSEL and quad-photodiode configuration where the 4 devices are on a 2 x 2 array on a dense 50-µm pitch. The transceiver module is completed by flip-chip soldering the Optochip onto a 8 mm x 8 mm high-speed high-density organic carrier. Optical access through the backside of the IC is provided through 2 optical vias. Electrical I/O is supplied through BGA pads on 0.8 mm pitch at the bottom of the module.High-speed characterization was carried out between 2 modules soldered to test cards, a transmitter (TX) and a receiver (RX) module. Each of the 4 optical outputs from the TX Optochip was coupled into a MMF and directed to individual photodiodes in the RX module. Eye-diagrams were measured for TX outputs as well as TX-to-RX links at data rates 20 Gb/s to 42 Gb/s. The 4 optical links operate error free up to 40 Gb/s, achieving a record data rate for multimode parallel optical transceivers.

No abstract

High performance computing (HPC), artificial intelligence (AI) and cognitive systems have initiated a new era of computing. Efficient thermal management technologies of these systems have been vital due to the increasing power density in the electronic components. In 2018 IBM delivered the fastest supercomputer of the world through Summit with 200 petaflops computing performance with LINPACK benchmarks. The system is both air and water cooled, where water is employed to cool the high power dissipated electronic components which are the IBM POWER9 processors and NVIDIA GPUs. In this paper, we highlight the overview of the thermal and mechanical design strategies applied on these systems. In air cooled systems, we discuss the fan and heat sink designs, as well as the preheating effect on PCI section. Liquid cooled system has a unique coldplate design which cool the processors and the GPUs with water. We examine the water flow path design for the processor and the GPUs by providing the thermal performance of the coldplate. Also, an overview of the cooling assemblies such as TIMs and air baffles in the servers are discussed. Moreover, unit and rack manifolds are investigated; flow and pressure distribution at the node and rack level are provided.

No abstract