ScyllaDB Benchmarks

NoSQL Database Benchmarks

Selecting a NoSQL database? How do you know which one is the best for your workload?

We constantly compare ScyllaDB performance vs. other NoSQL databases (throughput and latency). To help teams get a quick idea of how performance compares, we’re sharing those benchmarks here. We always recommend running your own assessment against your own workload and SLAs…and are available to help guide your own benchmarking.

Database Benchmark

ScyllaDB vs Apache Cassandra - Database Benchmark

  • ScyllaDB has 2x-5x better throughput and much better latencies than Cassandra
  • ScyllaDB adds and replaces nodes far faster than Cassandra
  • ScyllaDB finishes compaction 32x faster than Cassandra
  • A ScyllaDB cluster can be 10X smaller and 2.5X less expensive, yet still outperform Cassandra by 42%
Database Benchmark

ScyllaDB vs DynamoDB - Database Benchmark

  • ScyllaDB has 40.9X (provisioned) to 282.5X (on-demand) lower costs than DynamoDB
  • ScyllaDB has up to 4X better P99 latency than DynamoDB
  • ScyllaDB delivers single-digital ms latencies without an external cache
  • ScyllaDB does not have any per partition rate limit or throttling
Database Benchmark

ScyllaDB vs MongoDB – Database Benchmark

  • ScyllaDB has up to 20X higher throughput
  • ScyllaDB has up to 68X lower P99 latencies
  • ScyllaDB has up to 19X better price-performance
Database Benchmark

ScyllaDB vs Bigtable - Database Benchmark

  • ScyllaDB Cloud performs 26X better than Google Cloud Bigtable when applied with real-world, unoptimized data distribution
  • Google BigTable requires 10X as many nodes to accept the same workload as ScyllaDB Cloud
  • ScyllaDB Cloud was able to sustain 26x the throughput, and with read latencies 1/800th and write latencies less than 1/100th of Cloud Bigtable
Database Benchmark

ScyllaDB vs CockroachDB – Database Benchmark

  • Loading 10x the data into ScyllaDB took less than half the time it took for CockroachDB to load the much lesser dataset.
  • ScyllaDB handled 10x the amount of data.
  • ScyllaDB achieved 9.3x the throughput of CockroachDB at 1/4th the latency.
Database Benchmark

ScyllaDB vs Aerospike

  • ScyllaDB tested the same workload on equivalent hardware and achieved 40% to 100% greater throughput than Aerospike
  • A scaled-down ScyllaDB cluster should handle the same petabyte-scale workload for about $800K per year—roughly one-third less than Aerospike’s expected $1.23M annual cost.
  • Or, users could provision a ScyllaDB cluster at the exact same size and equivalent cost as the one proposed by Aerospike, yet get far greater throughputs with ScyllaDB.
  • ScyllaDB sustained  writes at 7.5 million inserts per second with 4 ms P99 latencies. While Aerospike maintained  ≤1ms P99 latencies, it provided significantly lower throughput than ScyllaDB.
Felipe Mendes Portrait

Felipe Mendes

Solutions Architect

1:1 Technical Strategy Session

Technical Strategy Session

Schedule a technical consultation with a ScyllaDB Solutions Architect.