8 vs. 12 Channel DDR5-6000 Memory Performance With AMD 5th Gen EPYC

Written by Michael Larabel in Memory on 20 November 2024 at 11:40 AM EST. Page 5 of 7. 29 Comments.
PostgreSQL benchmark with settings of Scaling Factor: 1000, Clients: 800, Mode: Read Only. 12c DDR5-6000 was the fastest.
PostgreSQL benchmark with settings of Scaling Factor: 1000, Clients: 800, Mode: Read Only, Average Latency. 8c DDR5-6000 was the fastest.
PostgreSQL benchmark with settings of Scaling Factor: 1000, Clients: 800, Mode: Read Write. 12c DDR5-6000 was the fastest.
PostgreSQL benchmark with settings of Scaling Factor: 1000, Clients: 800, Mode: Read Write, Average Latency. 12c DDR5-6000 was the fastest.

With the PostgreSQL database benchmarks using pgbench for this AMD 5th Gen EPYC 1P server configuration with Micron 7450 NVMe SSD storage, there wasn't any greater performance out of populating all twelve memory channels.

ClickHouse benchmark with settings of 100M Rows Hits Dataset, First Run / Cold Cache. 12c DDR5-6000 was the fastest.
ClickHouse benchmark with settings of 100M Rows Hits Dataset, Second Run. 12c DDR5-6000 was the fastest.
ClickHouse benchmark with settings of 100M Rows Hits Dataset, Third Run. 12c DDR5-6000 was the fastest.

The ClickHouse database benchmark though was able to enjoy a nice performance uplift out of the full twelve memory channel support with the AMD EPYC 9005 series.

Apache Cassandra benchmark with settings of Test: Writes. 12c DDR5-6000 was the fastest.

Apache Cassandra saw more limited gains going from eight to twelve DDR5-6000 memory modules with this AMD EPYC single socket server.

Related Articles