BLOG POST

Tech & Sourcing @ Morgan Lewis

TECHNOLOGY TRANSACTIONS, OUTSOURCING, AND COMMERCIAL CONTRACTS NEWS FOR LAWYERS AND SOURCING PROFESSIONALS

Increased Need for Cloud and Resulting Supply-Chain Challenges in Connection with Rapid Adoption of Gen AI

The rapid adoption of generative AI (gen AI) is driving increased cloud consumption.

This increased cloud consumption is largely due to the following:

  • Increased need for GPU power: Gen AI uses graphics processing units (GPU) power
  • AI development and use of models: AI developers need greater computing power to train gen AI models, and use of gen AI to create inferences and information requires robust computing power
  • Large enterprises: Large enterprises that have aging data infrastructures rely on the cloud for backup and resiliency
  • Backup GPU power: Redundancy and resiliency are critical due to the complexity of GPUs

Just as the COVID-19 pandemic pressure-tested the supply chain, this increased demand for cloud resources will do the same, and could have the following effects on the supply chain:

  • Increased demand for semiconductors (GPUs and tensor processing units) due to increased use and advancement of AI
  • Increased demand for hardware and components that intersect with AI applications and data centers (including power generation equipment)
  • Impact on pricing and availability due to increased competition in connection with demand for specific semiconductor chips
  • High demand for GPUs could impact the availability of other types of semiconductor chips
  • Strain on or expansion of industry due to such demand and competition
  • Temporary impact to other industries reliant on chips or hardware (e.g., the auto industry during the COVID-19 pandemic)

Based on the potential issues listed above, it is important for businesses to consider backup plans to help navigate the inevitable supply-chain disruptions and changes. Such plans may include:

  • Basing capacity on anticipated future needs
  • Considering long-term purchase agreements to guaranty manufacturing capacity and supply
  • Investing in data and operational supply-chain resiliency (e.g., forecasting and contingency planning)
  • Monitoring refresh cycles for related hardware that may be affected by such supply-chain issues