Why 400G in the Datacenter Requires Effective Thermoelectric Cooling

With consumer demand driving ever-faster, ever-better devices and networks, and cloud storage needs only increasing as well, the rollout of 400G in the datacenter has been much anticipated. Form factors and optical modules to support 400G are here, and expected to continue to launch throughout the year. In order to deliver the high performance needed to meet these data demands, laser packages will need to address cooling in an efficient and cost-effective way.

What’s Driving 400G?

Global mobile data traffic is increasing everyday, and a CAGR of 46% is expected through 2022. Advancements in cloud storage, edge computing, machine learning, artificial intelligence (AI) and machine-to-machine (M2M) communications all require nearly-instantaneous transfer of incredibly large amounts of data. Future technologies like next-gen virtual reality (VR) and autonomous cars will have huge data transmission needs as well as a need for a fully-connected IoT (Internet of Things) infrastructure.

There is a clear market demand for next-generation transceivers that support 400G data transfer. To remain viable from a business standpoint, however, these next-gen solutions need to offer low power consumption while performing at higher bandwidths to meet data transfer needs.

Why 400G Needs More Effective Thermoelectric Cooling

There are three main reasons why the rollout of 400G in datacenters will raise the need for cost-effective, efficient thermoelectric cooling:

• Higher speeds & longer reaches: The market demand isn’t just about increasing data rates. Communication links at FR4 (2km), as well as longer reaches including coherent (400G ZR at 80-120km), are becoming more and more common. This is particularly true for co-located datacenters. Faster data transfer rates and longer reaches put higher performance demands on the laser, and both applications will require cooling to maintain performance and avoid wavelength shift.
• Growing power consumption: QSFP-DD and OSFP transceivers are aiming to lower their power consumption to address MSA specifications. At the same time, they also need to drive performance to support the increasing download and transfer speeds. The manufacturer who strikes the perfect balance between performance and power usage will be positioned to win market share. In a highly standardized space, low-power, high-performance TECs (thermoelectric coolers) are one way to differentiate inside the package, where manufacturers have more control. The right TEC should help reduce total package cost and power consumption without sacrificing on performance.
• Shrinking form factors: This means that not only is heat density increasing, but component size is more critical than ever. Increasing heat density at the transceiver level is felt at the laser package level. We’ve seen an increase of approximately 3X over the last few years, with many packages now pumping 500mW or even 1W in just a few millimeters. This creates a need for small form factor TECs with better heat-pumping density performance for the size.

To keep up with growing consumer demand, transceiver manufacturers need to drive product innovation without increasing package cost in a highly cost-sensitive market. The right TEC solution can help you lower package power consumption and cost while maximizing performance, so your transceivers can meet 400G demands.

At Phononic, our engineers understand what you need to stay competitive. They can help you design a TEC configuration that’s optimized for your transceiver needs, including device size and processing temperature range. Learn more about designing TECs with the perfect balance between cost and performance with our Design Guide.