.webp)
Apple's entry-level laptop demonstrates competitive performance against cloud instances in industry-standard database benchmarks.
DuckDB's Gábor Szárnyas conducted a comprehensive benchmark comparison pitting Apple's entry-level MacBook Neo against cloud server instances to evaluate performance on demanding database workloads. The testing revealed surprising strengths for the 512GB laptop, particularly in cold-cache scenarios, though cloud infrastructure maintained advantages in memory-intensive operations.
The evaluation employed two industry standard benchmarks like ClickBench featuring 43 aggregations and filtering queries on a 100 million row dataset and TPC-DS comprising 24 tables and 99 complex queries incorporating window functions and advanced operations. The MacBook Neo faced competition from two cloud instances such as an AWS c6a.4xlarge with 16 AMD EPYC cores and 32GB RAM and a c8g.metal-48xl with 192 Graviton4 cores and 384GB RAM.
During ClickBench's cold-run test—measuring performance with empty caches—the MacBook Neo delivered exceptional results, completing all queries in under one minute and executing up to 2.8 times faster than both cloud instances. DuckDB attributed this performance advantage to the laptop's local NVMe SSD, which provides faster initial data access compared to network-attached storage on cloud platforms that dominated query runtimes.
The advantage reversed during hot-run testing, where cached data benefits larger systems. The c8g.metal-48xl completed the benchmark in 4.35 seconds, while the MacBook Neo finished at 54.27 seconds. However, the laptop remained competitive on median query runtimes against the mid-sized c6a.4xlarge instance, achieving performance only 13% slower despite the cloud instance having 10 additional CPU threads and 4 times more RAM.
TPC-DS benchmarking revealed the MacBook Neo's memory constraints at larger data scales. At SF100 scale, the laptop achieved a median query runtime of 1.63 seconds and total runtime of 15.5 minutes. At SF300, memory limitations became apparent, with the system utilizing up to 80GB of disk spillage and requiring 79 minutes to complete all queries, including a single query consuming 51 minutes.
Business Honor views the MacBook Neo's database performance as a strategic shift challenging traditional cloud infrastructure assumptions.
.webp)
