m8a.16xlargevsm8g.16xlarge

m8g.16xlarge is 26% cheaper than m8a.16xlarge on demand in US East (N. Virginia), saving $746.57 a month. They use different processor architectures (x86_64 and arm64), so software built for one may need rebuilding to run on the other.

At a glancem8a.16xlargem8g.16xlarge
On-demand$3.895/hr$2.872/hr26% less
Lowest rate$0.8559/hrSpot, can be interrupted$0.7971/hr7% lessSpot, can be interrupted
Size64 vCPUs, 256 GiB64 vCPUs, 256 GiB
ProcessorAMD EPYC 9R45 ProcessorAWS Graviton4 Processor
Architecturex86_64arm64

Prices in US East (N. Virginia)

“Off” is the discount against each type’s own on-demand price. Linux, USD per hour.

Optionm8a.16xlargem8g.16xlargem8g.16xlarge vs m8a.16xlarge
On-demandPay by the hour$3.895$2.872▼ 26%
SpotCan be interrupted78% off$0.855972% off$0.7971▼ 7%
1-yr reservedNo upfront—34% off$1.90—
3-yr reservedNo upfront—55% off$1.303—
1-yr savings planHourly commitment27% off$2.86227% off$2.11▼ 26%
3-yr savings planHourly commitment49% off$1.97549% off$1.457▼ 26%

Key differences

The biggest lead each one has.

Where m8a.16xlarge leads
Clock speedUp to 4.5 GHz vs 2.7 GHz
67% higher
IPv4 addresses per interface64 vs 50
28% higher
Where m8g.16xlarge leads
On-demand price$2.872 vs $3.895 per hour
26% cheaper

Specs

11 of 40 specs differ. Those are in bold.

Compute6 of 10 differm8a.16xlargem8g.16xlarge
vCPUs6464
Physical cores6464
Threads per core11
Architecturex86_64arm64
ProcessorAMD EPYC 9R45 ProcessorAWS Graviton4 Processor
CPU makerAMDAWS
Clock speedUp to 4.5 GHz2.7 GHz
Sustained clock4.5 GHz2.8 GHz
Processor featuresAMD Turbo; AVX; AVX2; AVX512–
BurstableNoNo
MemoryAll 1 matchm8a.16xlargem8g.16xlarge
Memory (GiB)256256
Networking3 of 8 differm8a.16xlargem8g.16xlarge
Network30000 Megabit30 Gigabit
Network baseline30 Gbps30 Gbps
Network peak30 Gbps30 Gbps
Enhanced networkingYes (ENA)Yes (ENA)
EFANoNo
IPv6YesYes
Max network interfaces1615
IPv4 addresses per interface6450
StorageAll 11 matchm8a.16xlargem8g.16xlarge
StorageEBS onlyEBS only
EBS optimizedYesYes
EBS over NVMeRequiredRequired
EBS throughput20000 Mbps20000 Mbps
EBS baseline bandwidth20,000 Mbps20,000 Mbps
EBS max bandwidth20,000 Mbps20,000 Mbps
EBS baseline throughput2,500 MB/s2,500 MB/s
EBS max throughput2,500 MB/s2,500 MB/s
EBS baseline IOPS80,00080,000
EBS max IOPS80,00080,000
Max EBS volumes4848
Instance details2 of 10 differm8a.16xlargem8g.16xlarge
Familym8am8g
CategoryGeneral purposeGeneral purpose
Current generationYesYes
Normalization factor128128
Bare metalNoNo
HypervisorNitroNitro
Placement groupsCluster, Partition, SpreadCluster, Partition, Spread
HibernationNoNo
Nitro EnclavesYesYes
Boot modesLegacy BIOS, UEFIUEFI

On-demand prices by region

Cheapest for m8a.16xlarge first. Linux, USD per hour.

Regionm8a.16xlargem8g.16xlargeDifference
US East (N. Virginia)Used in summaryus-east-1$3.895$2.872▼ 26%
US East (Ohio)us-east-2$3.895$2.872▼ 26%
US West (Oregon)us-west-2$3.895$2.872▼ 26%
Asia Pacific (Hyderabad)ap-south-2$4.098$2.053▼ 50%
Asia Pacific (Mumbai)ap-south-1$4.098$2.053▼ 50%
EU (Ireland)eu-west-1$4.341$3.201▼ 26%
Europe (Spain)eu-south-2$4.341$3.199▼ 26%
EU (Frankfurt)eu-central-1$4.666$3.441▼ 26%
AWS GovCloud (US-West)us-gov-west-1$4.909$3.62▼ 26%
Asia Pacific (Tokyo)ap-northeast-1$5.031$3.71▼ 26%

Keep exploring

Each type’s own page, and how it compares with others.

m8a.16xlargeGeneral purpose, m8a family
m8a.16xlarge prices and specsEvery size in the m8a family

Compare m8a.16xlarge with

m8i.16xlarger8a.16xlargec8a.16xlargem7a.16xlarge
m8g.16xlargeGeneral purpose, m8g family
m8g.16xlarge prices and specsEvery size in the m8g family

Compare m8g.16xlarge with

c8g.16xlargem8i.16xlarger8g.16xlargem8gd.16xlargem8gb.16xlargem8gn.16xlargem7g.16xlargem9g.16xlarge

Pay the committed rate, with commitment protection

The best committed rate on these two is 49–55% below on-demand. Connect your cloud accounts to unlock greater savings with commitment protection.

Connect AWS account

AWS EC2 Linux prices in USD. Monthly figures use 730 hours. Prices last updated 12 Nov 2026.