- Rust on arm64 completed CPU-intensive tasks up to 5x faster than x86
- Arm64 reduces cold start latency across all run times by up to 24%
- Python 3.11 on arm64 outperforms newer versions in memory-heavy workloads
Benchmarking of AWS Lambda this year shows that arm64 architecture consistently outperforms x86 across most workloads.
Tests covered CPU-intensive, memory-intensive, and light workloads across Node.js, Python, and Rust runtimes.
For CPU-bound tasks, Rust on arm64 completed SHA-256 hashing loops 4-5x faster than x86 Rust once architecture-specific assembly optimizations came into play.
Cold start and warm start efficiency
Python 3.11 on arm64 also outperformed newer Python versions, while Node.js 22 ran substantially faster than Node.js 20 on x86.
These results show that arm64 not only improves raw compute performance but also maintains consistency under varying memory configurations.
Cold start latency plays a crucial role in serverless applications, and arm64 delivers clear improvements over x86.
Across all runtimes, arm64 delivered 13–24% faster cold start initialization.
Rust, in particular, recorded nearly imperceptible cold start times at 16ms, which makes it well-suited for latency-sensitive applications.
Warm start performance also favored arm64, and memory-intensive workloads benefited from the architecture’s ability to handle larger memory allocations more efficiently.
Python and Node.js showed slightly more variability, although the gains from arm64 remained.
These performance improvements compound in production environments where frequent cold starts occur.
The cost analysis shows that arm64 delivers 30% lower compute costs on average compared to x86.
For memory-heavy workloads, cost savings reached up to 42%, particularly for Node.js and Rust.
Light workloads, which rely heavily on I/O latency rather than raw computation, showed minimal performance differences between architectures.
This shows that cost optimization matters more than runtime selection in these scenarios.
Across CPU-intensive and memory-intensive workloads, arm64 delivered stronger cost-to-performance ratios, confirming its value in production deployments.
These benchmarks indicate that arm64 should be the default CPU target for most Lambda workloads unless specific library compatibility issues appear.
Rust workloads on arm64 maximize both performance and cost savings, while Python 3.11 and Node.js 22 provide solid alternatives for other use cases.
Organizations that rely on Lambda for enterprise-scale applications or run multiple functions in a single data center will likely see clear efficiency improvements.
From a workstation perspective, the results suggest that developers compiling locally for CPU-intensive workloads may also benefit from arm64-native builds.
Although these benchmarks are extensive, individual workloads and dependency configurations can lead to different outcomes, so further testing is advisable before full-scale adoption.
Organizations leveraging Lambda for enterprise-scale applications or running multiple functions in a single data center will likely see tangible efficiency improvements.
From a workstation perspective, the results suggest that developers compiling locally for CPU-intensive workloads may also benefit from arm64-native builds.
Although these benchmarks are extensive, individual workloads and dependency configurations can yield different outcomes, so further testing is advisable before full-scale adoption.
Via Chris Ebert
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds. Make sure to click the Follow button!
And of course you can also follow TechRadar on TikTok for news, reviews, unboxings in video form, and get regular updates from us on WhatsApp too.
