Why Serverless Is the Future of Cloud Computing: Insights from Alibaba Cloud

Definition of Serverless

Serverless is a fully managed, API‑driven cloud service model that abstracts away underlying infrastructure. It comprises Function as a Service (FaaS) for compute and Backend as a Service (BaaS) such as messaging, object storage, and other cloud‑native services.

Evolution and maturity

The paradigm originated with AWS Lambda in 2014 and entered the Chinese market when Alibaba Cloud launched its serverless platform in 2017. According to a technology‑maturity curve, serverless has moved past the emergence and disillusionment phases and is now in a stable growth stage.

Stage 1 : High potential attracts broad attention.

Stage 2 : Early products reveal limitations, causing user frustration.

Stage 3 : Continued practice and product improvements lead to steady adoption.

Alibaba Cloud Serverless adoption

Since commercial launch in 2017, Alibaba Cloud’s Function Compute serves thousands of paid customers, hosts over 1 million functions, and processes more than 12 billion invocations per day. Core internal services such as Taobao, Fliggy, and Amap have migrated to this platform.

Key technical challenges

Vendor lock‑in concerns and the need for portable abstractions.

Fine‑grained resource scheduling and rapid scaling for bursty workloads.

Strong security isolation and multi‑tenant management.

Cultural shift from VM/container‑centric development to event‑driven serverless workflows.

Major technical upgrades (2020)

Architecture upgraded to next‑generation bare‑metal servers ("Shenlong") combined with secure containers ("Kangaroo"), delivering higher compute density, elasticity, and performance.

Cache layer introduced for accelerated container‑image distribution, enabling near‑second startup for gigabyte‑scale images.

Runtime rewritten for greater extensibility and reduced cold‑start latency.

Open‑source Serverless Devs

Alibaba Cloud released the Serverless Devs toolkit (open‑source on GitHub) to simplify development, debugging, deployment, and monitoring of serverless and containerized applications. The toolkit is modular: users can adopt any subset of its capabilities without fully migrating existing CI/CD pipelines, thereby reducing lock‑in risk.

Future opportunities and integration

Serverless is expanding beyond mini‑programs to cover e‑commerce peak traffic, video transcoding, AI inference, gaming, real‑time file processing, IoT data handling, and micro‑service architectures. Integration points include:

EventBridge event bus for cross‑service orchestration.

Third‑party SaaS platforms (e.g., DingTalk) that expose lightweight, customizable endpoints well‑suited to serverless execution.

Roadmap and expectations

Business scale is projected to grow more than threefold annually.

Product development will continue to focus on user pain points, improving tooling, observability, and lifecycle management.

Technical efforts will deepen compute, networking, caching, and runtime optimizations to achieve finer granularity, faster startup, and higher elasticity.