AWS r7g.16xlargevsAWS r7g.metal

r7g.16xlarge pairs 64 vCPUs with 512GB of RAM at $3.4272/hr On-Demand (about $2468/mo at 24×7). r7g.metal pairs 64 vCPUs with 512GB at $3.4272/hr (~$2468/mo). Both are priced identically per hour on-demand.

r7g.16xlarge (memory-optimized, AWS Graviton (ARM64)) and r7g.metal (memory-optimized, AWS Graviton (ARM64)) target different workload shapes. The vCPU:RAM ratios are different (64:512 vs 64:512), and depending on the architectures involved you may also see meaningful single-thread performance and per-hour-cost differences. This kind of cross-category comparison is most useful when you're early in the design phase and not yet sure whether your workload is CPU-bound, memory-bound, or balanced — once that's clear, the right pick is usually obvious.

On raw price-per-performance, the two are roughly tied on single-thread performance per dollar (880 vs 882 Sysbench points per $1/hr). That's the cleanest signal we have for "which one runs your workload faster per dollar," but it only matters if your workload is single-thread-bound; for parallel workloads the multi-core scores (192414 vs 192837) are what to weigh. Spot pricing flips many of these comparisons — when r7g.16xlarge drops to $0.8839/hr and r7g.metal drops to $0.9211/hr, the cheap-per-hour winner can swing meaningfully.

In practice, pick r7g.16xlarge when your workload is closer to memory-optimized (memory-bound work — in-memory databases, real-time analytics, large caches). Pick r7g.metal when it's closer to memory-optimized (memory-bound work — in-memory databases, real-time analytics, large caches). When neither side is obviously right, the cheaper hourly rate usually wins for fault-tolerant batch workloads, while the higher single-core score usually wins for latency-sensitive web traffic. The regional pricing tables linked from each instance page below show where each is currently cheapest — sometimes a >20% regional gap flips the comparison entirely.