If One Thing Is Clear from MWC 2022, It Is That Operators Need Smarter, as Well as More, Data Centers

Subscribe To Download This Insight

By Reece Hayden | 1Q 2022 | IN-6481

Building new data center (DC) sites is no longer enough to keep in lockstep with digital transformation led processing demand, instead operators and enterprises must augment their global data center (DC) strategy by introducing artificial intelligence (AI) to improve operations and management (O&M). Creating ‘smarter’ DCs was a critical topic at Mobile World Conference (MWC) 2022 with Nokia and Huawei both heralding the virtues of AI-managed DC solutions. But this drive towards a ‘smarter’ solution was in the works way before MWC’22, as NVIDIA and Equinix announced their partnered ‘Nvidia Launchpad’ in Mid-2021.

Registered users can unlock up to five pieces of premium content each month.

Log in or register to unlock this Insight.

 

Vendors Target Life Cycle O&M

NEWS


Enhancing environmental and operational efficiency were of chief concern for operators at MWC 2022, and sone of the primary targets were, understandably, DC sites. Big-named players, such as Huawei and Nokia underscored the commercial and environmental benefits of ‘smart’ AI managed DCs. As Huawei drew attention to its prefabricated, pre-validated data centers that deliver a variety of benefits across the entire life cycle (including power, construction, and cooling efficiency), including a reduction in O&M costs by 35%, through AI-driven life cycle management. Whilst Nokia announced its ‘Edge Network Controller’, as part of its ‘Adaptive Cloud Networking’ solution. This service looks to automate lifecycle management of edge cloud locations, to ensure sites can scale dynamically to meet 5G-spurred DC/edge demand. It seeks to invigorate telco cloud design and deliver value across the 5G digital network value chain.

However, Nvidia were the first movers, when mid-way through 2021, the AI-leader announced their 'Nvidia Launchpad’ product, in association with Equinix; a software enabled AI platform offering enterprises, utilizing colocation facilities, the opportunity to deploy and test AI technologies ‘as-a-service’ within DCs, without re-building servers or incurring long-run costs for solutions that may not be operationally beneficial.

'Smarter' O&M Will be Necessary to Hitting DC Commercial Value Targets

IMPACT


ABI Research has previously evaluated the environmental and social benefits of AI-driven DCs (“Greener” Data Centers in an Increasingly Digital and Data-Driven World (IN-6387)), but it is evident that AI-driven O&M will also stimulate commercial opportunities by mitigating the exogenous constraints (environmental, legal, access to power, limited human capital) that a changing economy will impose on DC operators/enterprises.

AI-enabled data and cloud O&M will be a critical piece in the jigsaw that is data center efficiency. Digital transformation and 5G-enabled technologies, including IoT, will contribute to a sharp rise in data transmission, placing additional strain on access to network work. Traditional strategies that rely on increasing the global distribution of DC locations will be too slow to meet global demand, because exogenous constraints  increase the time-to-market. Instead, ABI Research suggests that DC efficiency led by AI O&M strategies will be the only viable approach that ensures that operators and enterprises can keep in lockstep with demand, whilst driving sustained commercial value. The benefits of AI-driven O&M are varied, but importantly, it improves DC performance for operators, enterprises, and service providers as clarified below:

  • DC Operators
    • Reduce human capital requirements, as centralized AI operations can manage multiple servers.
    • AI threat detection can ensure that DC uptime is maximized.
    • AI can sense real-time data demand changes and dynamically scale operations to match.
    • Ensure telco cloud/edge capacity can scale to meet low latency, dynamic performance requirements.
  • Enterprises
    • ‘Nvidia Launchpad’ is provided ‘as-a-service’ and offers ‘test’ periods to ensure that commercial value is targeted.
    • Abstracting hardware from DCs will increase time-to-market and reduce cost of integration.
  • Service providers
    • Combine Nokia’s ‘Adaptive Cloud Networking’ with edge network slicing to ensure service providers can minimize network cost savings for enterprise.
    • Leverage 5G slicing to ensure end-to-end managed network infrastructure.

It is evident that both operators and enterprises can operationally and commercially benefit from ‘smarter’ AI-driven DCs. For this reason, ABI Research recommends that all players, regardless of size or global position, should explore how augmenting DC strategy with AI could improve long-run commercial and operational value. If players begin to explore these solutions, the DC market will begin to rapidly adapt over the next 5 years, but how do we see the market changing?

DC Market Will Look Very Different in a Few Years

RECOMMENDATIONS


One expectation is that DC facility construction will slow, however ABI Research does not expect growth to drop off soon, even with the introduction of AI-driven solutions. It is likely that top operators will continue to expand their global footprint over the next few years to meet global networking demands, especially in emerging regions such as APAC, South America, and Middle East/Africa. Moreover, due to the infrastructure requirements for AI solutions, operators will be forced to reimagine their existing DC solutions. This will, at least in the short run, slow consolidation, because tailor-made facilities will be necessary to deliver ‘smarter’ AI-driven facilities.

It will be interesting to see how AI-driven O&M will disseminate through the DC market. ABI Research expects that all DC operators, and enterprises requiring third-party services, can benefit from AI-driven O&M. But, for enterprises to receive the full benefits that this technology offers, colocation facilities must be ready with mature offerings that can adequately support widespread server-based AI deployment. These colocation providers must ensure that they:

  • Offer more scalable space that can expand to accommodate AI-driven increase in data processing.
  • Enable availability of high-power servers.
  • Provide network peering to allow third parties to align with multiple carriers.
  • Build infrastructure that aligns with AI requirements

Looking at the future of the DC market, ABI Research recommends that operators must follow three concurrent strategies to optimize operational efficiency and drive commercial value: (1) complete global DC coverage; (2) expand existing data center facility footprints; (3) optimize infrastructure to utilize and enable AI-driven solutions. MWC 2022’s announcements, although not innovative, have bought into sharp focus the requirement for operators and enterprises to reassess their strategy and look at AI-integration and a shift towards site efficiency. For ABI Research, early adopters of this new three-pronged approach will, over the next 10 years, become market leaders, as they are able to mitigate pressure from legal, environmental, and construction bottleneck constraints that are increasingly becoming more globally prevalent.

 

Services