AWS Lambda timeout increase to 90 minutes targets media transcoding

The extension of execution limits directly addresses a long-standing bottleneck for serverless media processing. By allowing 90-minute runtimes, Amazon Web Services enables engineers to handle high-bitrate or long-form video transcoding within a single function, eliminating the need to shard files or manage complex state machines via Step Functions. This shift makes serverless architectures more viable for heavy lifting in the streaming supply chain, potentially reducing the overhead of maintaining dedicated EC2 clusters for batch processing. Watch for whether competitors like Google Cloud Functions or Azure Functions match this duration to prevent developer migration toward Lambda Managed Instances for compute-heavy video workflows.

AWS Lambda's extended timeout positions it against competing serverless platforms that have already pushed execution limits for compute-heavy tasks. Google Cloud Functions has supported up to 60-minute timeouts for second-generation functions since 2023, while Azure Functions Premium plan allows unbounded execution on dedicated infrastructure. The 90-minute ceiling on Lambda Managed Instances now exceeds both rivals for single-invocation duration, giving AWS a concrete advantage for long-form video processing pipelines that previously required orchestration layers. Ericsson launched its AI in RAN commercial software subscription on June 11th, claiming up to 20% higher downlink throughput and up to 10% better spectral efficiency across more than 15 live deployments using existing baseband silicon, illustrating how infrastructure vendors across sectors are extending execution windows to accommodate AI-driven workloads without architectural fragmentation. The business case for longer serverless timeouts aligns with broader cloud infrastructure spending trends in media and entertainment. Nokia and AWS have recently showed the industry's first agent for network slicing with du and Orange in March, and the world's first commercial 5G service on a cloud-hosted SaaS core with Citymesh in Belgium in February, demonstrating AWS's expanding footprint in telecom and media infrastructure beyond traditional compute. Nokia reckons its autonomous networks portfolio is delivering measurable results already, with operators achieving automation rates higher than 90 percent and service delivery times of four hours or less. AWS Lambda's timeout extension fits this pattern of cloud providers absorbing workloads that previously required dedicated hardware or complex orchestration, reducing total cost of ownership for streaming operators managing transcoding at scale. Technical benchmarks from the broader AI-RAN and network automation space provide useful context for understanding why execution duration matters in production environments. Nokia's platform combines its anyRAN software with Nvidia's Aerial AI-RAN computing environment and can be deployed three ways, with spectral efficiency gains of more than 20 percent already delivered and a roadmap targeting 50 percent by 2027. These gains require sustained compute sessions that mirror the demands of multi-hour transcoding jobs. Ericsson is making the opposite bet, unveiling AI-RAN products built entirely on custom silicon rather than Nvidia GPUs ahead of Mobile World Congress 2026, arguing that neural-network accelerators can be added directly to beamforming chips. T-Mobile has trialed Ericsson's AI-driven link adaptation software on its live 5G Advanced network, reporting gains in spectral efficiency. The parallel between these sustained AI inference workloads and media transcoding underscores why AWS chose 90 minutes as the new ceiling: it covers the typical duration of a 4K feature-length encode while remaining within serverless billing models.