Architecture & Sizing
RAM-to-vCPU Sizing Guide: C7g vs M7g vs R7g
Choosing between 2:1, 4:1, and 8:1 RAM-to-vCPU ratios on Graviton3
Understanding RAM ratios is essential for right-sizing container nodes and virtual machines. Compare the identical Graviton3 CPU core paired with different memory capacities.
1. Architectural & Silicon Specs Head-to-Head
Comparing underlying processor microarchitectures, memory ratios, and Nitro generation features using representative xlarge sizing.
c7gARM64 (Graviton)
C7G Compute Optimized
ProcessorAWS Graviton3 Processor
Clock Speed2.5 GHz
RAM / vCPU2.0 GiB/vCPU
Max Network30 Gbps
Max EBS IOPS80,000
AWS NitroSupported
Min Rate (c7g.xlarge)$0.0982/hr
Monthly Est.$71.69/mo
Cost per vCPU$0.0245/hr
Cost per GiB RAM$0.0123/hr
CoreMark Benchmark99,782
CoreMark / $1,016,110
Price IndexLowest Cost
Save 35% vs most expensive
Compute / $ Value100% (Leader)
m7gARM64 (Graviton)
M7G General Purpose
ProcessorAWS Graviton3 Processor
Clock Speed2.5 GHz
RAM / vCPU4.0 GiB/vCPU
Max Network30 Gbps
Max EBS IOPS80,000
AWS NitroSupported
Min Rate (m7g.xlarge)$0.1166/hr
Monthly Est.$85.12/mo
Cost per vCPU$0.0291/hr
Cost per GiB RAM$0.0073/hr
CoreMark Benchmark103,066
CoreMark / $883,928
Price Index78% of highest
Save 22% vs most expensive
Compute / $ Value87% of leader
r7gARM64 (Graviton)
R7G Memory Optimized
ProcessorAWS Graviton3 Processor
Clock Speed2.5 GHz
RAM / vCPU8.0 GiB/vCPU
Max Network30 Gbps
Max EBS IOPS80,000
AWS NitroSupported
Min Rate (r7g.xlarge)$0.1502/hr
Monthly Est.$109.65/mo
Cost per vCPU$0.0376/hr
Cost per GiB RAM$0.0047/hr
CoreMark Benchmark102,571
CoreMark / $682,896
Price Index100% of highest
Compute / $ Value67% of leader
2. Side-by-Side Sizing & Pricing Matrix
Direct price and benchmark comparison across matching instance tiers in the lowest-cost region. Lowest hourly cost in each row is highlighted in green.
| Size Tier | vCPU | Memory | c7gLinux Rate | m7gLinux Rate | r7gLinux Rate |
|---|---|---|---|---|---|
| .medium | 1 | 2 GB | |||
| .large | 2 | 4 GB | |||
| .xlarge | 4 | 8 GB | |||
| .2xlarge | 8 | 16 GB | |||
| .4xlarge | 16 | 32 GB | |||
| .8xlarge | 32 | 64 GB | |||
| .12xlarge | 48 | 96 GB | |||
| .16xlarge | 64 | 128 GB |
3. Workload Decision Framework
Clear, actionable rules on when each family wins based on workload profile, runtime language, and licensing boundaries.
Recommended:C7g (2 GiB / vCPU)
When: CPU-bound APIs, batch jobs, video encoders, and Go/Rust microservices
Why: Pay only for the memory you consume; avoid paying for idle RAM.
Recommended:M7g (4 GiB / vCPU)
When: Standard web apps, Ruby on Rails, Django, and Node.js with moderate heap usage
Why: Balanced headroom to prevent Out-Of-Memory (OOM) kills during traffic bursts.
Recommended:R7g (8 GiB / vCPU)
When: JVM applications with large heaps, Redis caches, and database servers
Why: High RAM capacity per core without paying for unused vCPUs.
5. Related Architecture Comparisons
Explore adjacent generational comparisons and silicon trade-offs.