Fresh off the quarterly earnings for Microsoft, Google, and Amazon, we are releasing the Q3 2024 Cloud Cost Report, an analysis of cloud usage based on anonymized Vantage customer usage. Vantage is a cloud cost visibility and optimization platform, with a unique view into industry trends, thanks to tens of thousands of connected infrastructure accounts across 16 cloud providers. To discuss this report in more detail, join our growing Slack Community of over 1,000 engineering leaders, FinOps professionals, and CFOs. View past reports here.
This year, there was speculation that large companies were beginning to transition entirely to on-prem environments. Despite this, data shows all three cloud providers continuing to increase their earnings by billions of dollars.
Azure and GCP, in particular, are increasing market share at a higher percentage than AWS. Organizations are increasingly adopting distributed architectures, indicating they care more about using the right tool for their needs. This often results in multi-cloud or hybrid-cloud setups.
We’ve been analyzing the distribution of processor usage across EC2 and other AWS services. Findings reveal that services like RDS, ElastiCache, and OpenSearch predominantly use Graviton processors. This can be attributed to Graviton’s cost and performance benefits.
On the other hand, for EC2, Intel processors make up a majority of overall spend. This could be attributed to users' reluctance to migrate legacy workloads or to Intel's broader range of instances, which have more expensive options (e.g., u-24tb1.112xlarge).
However, as you can see for newer generations, Graviton is gaining a larger portion of the EC2 share. Again, because of cost and performance benefits, and also, for newer generations, AWS has a strategy of releasing Graviton instances first, making it the only choice for early adopters.
Last quarter saw a steep decline in On-Demand spend after increases in previous quarters. This quarter is a continuation of that, as EC2 users continue opting for Reserved Instances and Savings plans to commit and save.
Over the last few quarters, we've seen increased experimentation with GPU instances. This quarter supports the findings that companies are incorporating their GPU workloads into their commitments for more cost savings.
Datadog users, on the other hand, are not taking advantage of commitments. Throughout the year, On-Demand spend for APM has remained between 82-87% without much fluctuation.
One possible reason users are avoiding commitments could be the need to engage with account managers and negotiate non-public rates for minimum commitments. However, using commitments could save you an estimated 20-50% on costs. It’s recommended to start conservatively, as you can always increase your commitments over time.
Together, Log Management and APM account for 40% of Datadog spend. The data shows both of which have untapped, low-effort opportunities for savings. Last section, we discussed how APM users could save by utilizing commitments.
Similarly, in last quarter's report, we analyzed Log Management classes and recommended reducing retention durations as an effective way to lower costs.