
Datadog recently released interesting data about their customers’ AWS usage and spending patterns. Here are some of the key findings, along with Wingspan’s perspectives:
1. AI Usage
Data Point: 74% of GPU usage is spent on the least expensive GPU types, indicating that most customers are still in the experimentation phase of AI adoption.
Wingspan Opinion: Your portfolio companies should, at the very least, be engaging in AI experimentation. This phase is critical for gaining competitive insights and building capabilities for the future.
2. ARM Adoption
Data Point: There is a steady shift toward ARM processors due to their superior energy efficiency and cost-effectiveness.
Wingspan Opinion: Transitioning applications to an ARM architecture isn’t straightforward. Companies must evaluate use cases, compatibility issues, and whether the potential benefits justify the effort. Often, application refactoring is required for a successful transition.
3. Containers
Data Point: One-third of EC2 compute spend is allocated to containers, yet over 80% of this spend is wasted on idle resources.
Wingspan Opinion: To combat waste, proactively analyze application resource requirements and employ strategies to minimize idle time. This includes:
Identifying and shutting down idle resources.
Implementing auto-scaling to adjust resources based on demand.
Scheduling resources for non-peak times.
Right-sizing instances to match workloads.
Leveraging reserved or spot instances for predictable and flexible workloads, respectively.
4. Previous-Generation Technology Usage
Data Point: 83% of organizations still rely on previous-generation EC2 instance types.
Wingspan Opinion: Modernization efforts are essential to unlock cost savings and improved capabilities. However, balance is key—adopt reliable, newer technologies without rushing to the bleeding edge, which can introduce risks.
5. Data Transfer Costs
Data Point: Traffic across availability zones accounts for half of data transfer costs.
Wingspan Opinion: High availability across multiple availability zones is often necessary, but it’s not always justified. Evaluate your specific needs and aim for a pragmatic approach to reduce unnecessary costs.
Conclusion
Your portfolio companies are either already in the cloud or transitioning to it. Significant efficiencies can be achieved through the right combination of:
Effective financial audit and governance processes.
Application modernization initiatives.
At Wingspan, we’ve observed that most portfolio companies lack comprehensive skills in both architecture and FinOps. This is where we can help. Our services include:
Cloud strategy development.
Modern cloud architecture design.
Cloud migration planning and execution.
Cloud financial operations (FinOps).
Negotiations with hyperscalers to optimize agreements.
About Wingspan Equity
Wingspan Equity is a Private Equity secondary-investment and advisory firm dedicated to accelerating growth and creating lasting value for technology organizations held by private equity. We go beyond capital, offering:
Operational expertise: Proprietary mechanisms and accelerators to drive performance.
Alignment of goals: A co-investment model that ensures all parties focus on shared outcomes.
For more insights, visit the original Datadog article: State of Cloud Costs.
Comments