Though 400 Gigabit Ethernet (400G) is still working its way into production networks, 800G demand from datacenters has already arrived.
Transceiver and cable vendors and semiconductor companies are racing to develop and test 800G components at the lower layers of the protocol stack. But they are finding fragmented standards development and timelines are resulting in complexity and confusion along the way.
Unfortunately, stakeholders do not have the luxury of waiting for these challenges to be sorted. They must move ahead now to meet critical market needs.
Doing so with confidence means leaning on testing earlier and more comprehensively than any previous evolution required. Understanding what can go wrong—and when—is the first step toward plotting a strategy for early-stage success.
Explore the dynamics and early activity defining this nascent market in our latest whitepaper .
Next-gen network pursuits reveal new challenges
Although 800G is being defined based on 400G technology, there are significant new technical challenges at all layers. These challenges impact the entire ecosystem, from chipset and transceiver/cable vendors to NEMs and hyperscalers.
From our early testing, several key challenges are materializing:
Evolving standards mean constant change. Standards for 800G PCS (Physical Coding Sublayer) and adaptation of 400G FEC (Forward Error Correction) for 800G continue to evolve, bringing small changes but also some major leaps, such as doubling the PAM (Pulse Amplitude Modulation) speeds.
Early devices not supporting complete standards. Auto-Negotiation (AN) and Link Training (LT) for electrical signal transmission can increase link connection issues and will need to be addressed as devices mature in alignment with defined standards.
More power means more heat. 800G optics generate high levels of power and heat, impacting performance if not cooled efficiently.
Little room for error. The jump to 800G doubles the sampling speed and the symbol rate, so what were minor issues at 400G are impacting electrical performance at 800G.
There remains so much to be sorted. But there is also good news: FEC statistics are emerging as a highly valuable analysis tool for quickly and proactively assessing physical layer health.
With standards still being determined, components scarce and expensive due to supply chain issues, and new complexity introduced, proper testing is the single-best tool stakeholders have to plot a strategic path forward.
Safely accelerating the journey to 800G
Early testing work has demonstrated the viability of next generation high-speed Ethernet devices and readiness to support initial demands.
Already, Spirent has helped validate 800G optical transceiver interoperability on test links that emulated hyperscale data center traffic. We have also validated 800G optical transceiver performance via real-world traffic emulation to support continued expansion of online applications and services in data centers and partnered to make strides in providing vendor-neutral .
Dive deeper into 800G’s challenges and what leading stakeholders are doing to advance in this fast-moving market. Download our whitepaper today.