The growth of data centers is a constant theme in the trade press. Typically, this growth is presented as an encouraging trend, creating demand for a variety of systems and components. But for data center managers hyperscale data centers are mostly cost centers; necessary evils that require cost reduction and cost analysis, and do not produce profits for their owners.
All this has consequences for optical component manufacturers. These implications have both positive and negative implications for the component sector. On the one hand, the cost sensitivity of data centers points towards the need for plain vanilla standardized components that are hard to distinguish in the marketplace. On the other hand – and somewhat paradoxically – CIR believes that it also indicates a new and potentially profitable direction for components companies in the form of what we shall call “smart optical components.”
Cost Cutting and Commoditization
In a recent paper, Gartner recommended that to lower data center costs one must “standardize technologies. Gartner also stressed that cost reduction implied a need to “centralize and consolidate.”
A technology standardization imperative, as suggested by Gartner, could be treated as a call to be suspicious of novel technologies. And while “centralize and consolidate” appears to be a call for the latest high data rate technologies – 100 Gbps and 400 Gbps – Gartner also seems to be saying that high bandwidth interconnect deployments should be carried out adopting a strategy that emphasizes the minimization of technology risks.
However, a strong focus on technology standardization makes it harder to distinguish products in the marketplace. And at the current level of technology evolution, many likely technology innovations in optical components are quite likely to occur in the 100 Gbps to 400 Gbps range where we are supposed to be avoiding technology risk, according to Gartner.
On face of it then, Gartner’s pronouncements aimed at data center managers don’t sound like good news for components makers. Nonetheless, CIR does not see optical components being quickly pushed toward commoditization under the heel of the cost cutting policies of data center managers. Indeed, we think a different aspect of these policies will bring a new kind of opportunity to the optical components manufacturers.
Smart Components as a Coming Opportunity
The starting point here is the same Gartner article that we cited before, which not only stresses the need for gross cost cutting but also for the fine tuning of cost monitoring. More specifically, Gartner says that data center managers should – as their “first priority” make all IT-related activity costs transparent. Gartner also insists that data center manager’s “benchmark how their organization’s spending compares to your peers”
To CIR this should be taken to mean that monitoring of and data collecting from the data center is an even higher priority than standardization and network consolidation. This Gartner pronouncement seems to indicate the need for intelligence in the network and this intelligence could be provided at either the equipment or the component level – in the latter case we are saying that there will be a growing opportunity for what CIR is calling “smart components.”
What CIR expects is that the cost cutting concerns of data center managers will boost demand for smart optical components in a way that gets around the commoditizing impact of the other cost cutting strategies mentioned above. The balance of this article reviews areas where, it seems to us, there are opportunities for smart optical components that fit in well with the current needs of data center managers.
Smart Components as a Cost Cutting Strategy
Incorporating intelligence into optical components is not a new idea, performance-monitoring devices have been integrated with components for some time, but with a different goal. What CIR specifically envisions here is a class of optical components with integrated optical monitoring that is tuned specifically to cost collection and analysis needs.
Think of smart optical components as sensor devices in a specialized industrial Internet-of-Things (IoT). For example, integrated optical monitors might provide bit-error rate (BER) data or information on the presence of a signal on a given optical channel. They would be designed so that they not only provided guidance on the operational level of the data center, but also what inefficiencies were costing the data center owners:
- This cost collection functionality would be a new direction for optical components, although it could be built on the back of the kind of operational optical performance that exists now. This mission could be completed in other ways (at the equipment level, for example). But what component level cost monitoring can bring to the table is provide cost monitoring at numerous locations throughout the data center network, thereby mimicking the kind of thing that other – more familiar – IoTs do.
- Integration technology would be key to creating a class of smart optical components of the kind we are talking about here. Adding cost monitoring capabilities to optical components would be self-defeating if it is itself very expensive. Getting smart optical components to a point where they are cost effective in a data center context, CIR suspects, will be a challenge. So far optical monitoring is associated primarily with long-haul networks, where budgets are significantly higher than in data centers. However, we think that in this area, optical component firms can draw comfort from the integration achievements embodied in optical engines and – more broadly – what has been done with silicon photonics.
Finally, the technical challenges for smart optical components will not be the only technical. Getting data center managers to buy into very new ideas are always hard, since their careers can depend on their technology choices. These data center managers do not buy the components themselves, of course, but they will or won’t “vote” for smart optical components depending on what equipment they buy.