m7g.2xlarge
General purpose · AWS Graviton3 (Arm)
PostgreSQL benchmark results for the m7g.2xlarge AWS EC2 instance with 8 vCPU, 32 GiB RAM, and arm64 architecture.
Compute
- vCPU
- 8
- Memory
- 32 GiB
- Mem / vCPU
- 4 : 1
Platform
- Family
- m7g
- Architecture
- arm64
- Performance
- Fixed
Cost & region
- On-demand
- $0.3264/hr~$238/mo
- Region
- us-east-1
Chart
Change the metric and disk to see how the numbers shift as the dataset grows — and where it stops fitting in memory.
Open markers sit past the RAM limit (~24 GB of 32 GiB): the dataset no longer fits in cache, so the working set spills to disk and throughput drops.
All results
6 of 6 rows
| 1 GB | gp3-baseline | 77,197±1.6% | 221,629 | 0.41±1.6% | 1.81±0.2% | 2.22±0.2% | 9,650 | 2,412 |
| 10 GB | gp3-baseline | 75,741±0.8% | 217,449 | 0.42±0.8% | 1.82±0.3% | 2.23±0.2% | 9,468 | 2,367 |
| 10 GB | gp3-max | 67,283±0.4% | 138,640 | 0.48±0.4% | 1.89±0.1% | 2.28±0.1% | 8,410 | 2,103 |
| 1 GB | gp3-max | 67,275±0.2% | 138,625 | 0.48±0.2% | 1.89±0.2% | 2.29±0.3% | 8,409 | 2,102 |
| 50 GB | gp3-max | 26,835±1.2% | 55,295 | 1.19±1.2% | 2.89±1.3% | 3.78±1.4% | 3,354 | 839 |
| 50 GB | gp3-baseline | 5,267±0.7% | 15,121 | 6.08±0.7% | 18.48±10.0% | 27.52±23.6% | 658 | 165 |