Silicon Triangle

AWS Burstable Comparison: T4g vs T3 vs T3a

Low-cost baseline compute with CPU credit burst mechanics

Burstable T-series instances power staging environments, developer boxes, low-traffic APIs, and background job consumers. Compare CPU credit accumulation, price-performance, and burst capabilities.

1. Architectural & Silicon Specs Head-to-Head

Comparing underlying processor microarchitectures, memory ratios, and Nitro generation features using representative large sizing.

t4gARM64 (Graviton)

T4G General Purpose

ProcessorAWS Graviton2 Processor

Clock Speed2.5 GHz

RAM / vCPU4.0 GiB/vCPU

Max NetworkUp to 5 Gbps

Max EBS IOPS15,700

AWS NitroSupported

Min Rate (t4g.large)$0.0448/hr

Monthly Est.$32.70/mo

Cost per vCPU$0.0224/hr

Cost per GiB RAM$0.0056/hr

CoreMark Benchmark39,596

CoreMark / $883,839

Price IndexLowest Cost

Save 46% vs most expensive

Compute / $ Value100% (Leader)

Browse all 7 t4g sizes →

t3x86_64

T3 General Purpose

ProcessorIntel Skylake E5 2686 v5

Clock Speed3.1 GHz

RAM / vCPU4.0 GiB/vCPU

Max NetworkUp to 5 Gbps

Max EBS IOPS15,700

AWS NitroSupported

Min Rate (t3.large)$0.0832/hr

Monthly Est.$60.74/mo

Cost per vCPU$0.0416/hr

Cost per GiB RAM$0.0104/hr

CoreMark Benchmark27,293

CoreMark / $328,041

Price Index100% of highest

Compute / $ Value37% of leader

Browse all 7 t3 sizes →

t3ax86_64

T3A General Purpose

ProcessorAMD EPYC 7571

Clock Speed2.5 GHz

RAM / vCPU4.0 GiB/vCPU

Max NetworkUp to 5 Gbps

Max EBS IOPS15,700

AWS NitroSupported

Min Rate (t3a.large)$0.0493/hr

Monthly Est.$35.99/mo

Cost per vCPU$0.0246/hr

Cost per GiB RAM$0.0062/hr

CoreMark Benchmark25,986

CoreMark / $527,099

Price Index59% of highest

Save 41% vs most expensive

Compute / $ Value60% of leader

Browse all 7 t3a sizes →

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	t4gLinux Rate	t3Linux Rate	t3aLinux Rate
.nano	2	1 GB	$0.0028/hr $2.0/moLowest Cost	$0.0052/hr $3.8/mo+86%	$0.0031/hr $2.3/mo+11%
.micro	2	1 GB	$0.0056/hr $4.1/moLowest Cost	$0.0104/hr $7.6/mo+86%	$0.0062/hr $4.5/mo+11%
.small	2	2 GB	$0.0112/hr $8.2/moLowest Cost	$0.0208/hr $15.2/mo+86%	$0.0123/hr $9.0/mo+10%
.medium	2	4 GB	$0.0224/hr $16.4/moLowest Cost	$0.0416/hr $30.4/mo+86%	$0.0246/hr $18.0/mo+10%
.large	2	8 GB	$0.0448/hr $32.7/moLowest Cost	$0.0832/hr $60.7/mo+86%	$0.0493/hr $36.0/mo+10%
.xlarge	4	16 GB	$0.0896/hr $65.4/moLowest Cost	$0.1664/hr $121.5/mo+86%	$0.0986/hr $72.0/mo+10%
.2xlarge	8	32 GB	$0.1792/hr $130.8/moLowest Cost	$0.3328/hr $242.9/mo+86%	$0.1971/hr $143.9/mo+10%

3. Workload Decision Framework

Clear, actionable rules on when each family wins based on workload profile, runtime language, and licensing boundaries.

Recommended:T4g (Graviton2)

When: Development environments, staging clusters, and microservice sidecars

Why: Lowest hourly price of any AWS compute tier with generous baseline CPU credit accumulation.

Recommended:T3a (AMD)

When: x86_64 development or tools that require standard x86 instruction sets

Why: 10% cheaper than Intel T3 with identical CPU credit allocation rules.

Recommended:T3 (Intel)

When: Workloads requiring Intel QuickSync or specific Intel hardware features

Why: Intel Skylake baseline with Unlimited mode support for burst credit borrowing.

4. Architecture & Migration Considerations

Technical caveats, instruction set differences, and compiler flags to know before deploying.

T4g Unlimited by Default

T4g instances run in 'Unlimited' mode by default. If your application sustains 100% CPU beyond earned credits, surcharges apply ($0.04/vCPU-hour). For steady high CPU, switch to C7g/M7g.

5. Related Architecture Comparisons

Explore adjacent generational comparisons and silicon trade-offs.

1. Architectural & Silicon Specs Head-to-Head

2. Side-by-Side Sizing & Pricing Matrix

3. Workload Decision Framework

4. Architecture & Migration Considerations

T4g Unlimited by Default

5. Related Architecture Comparisons

T2 vs T3 vs T4g