a2-megagpu-16g

    $84.5426/hr

    ~$61716.08/month on-demand in us-south1

    The a2-megagpu-16g is an Accelerator optimized instance with 96 vCPUs and 1360 GiB of memory, 16x NVIDIA nvidia-tesla-a100 with up to 32 Gbps bandwidth, starting at $84.543 per hour or $76.412 per hour with spot instances.

    Last updated: Jan 16, 2026, 12:03 PM UTC

    Pricing Options

    CUD

    Committed Use Discounts

    1 Year
    $80.7013/hr
    ~$58911.97/mo
    Save 5%
    3 Year
    $77.7957/hr
    ~$56790.89/mo
    Save 8%

    Spot Pricing

    Up to 90% savings - flexible workloads

    Spot Price
    $76.4121/hr
    ~$55780.81/mo
    Save 10%

    Instance Specifications

    Detailed hardware and capabilities for a2-megagpu-16g

    Processor

    Architecturex86_64
    vCPUs96
    Memory1360 GiB
    Memory per vCPU14.17 GiB

    Network

    PerformanceUp to 32 Gbps
    Bandwidth32 Gbps
    Max Interfaces8
    Enhanced Networking

    Storage

    Persistent Disk Only

    GPU

    Count16
    ManufacturerNVIDIA
    Modelnvidia-tesla-a100
    Memory40 GiB
    Total GPU Memory640 GiB

    Capabilities

    Familya2-megagpu
    GenerationCurrent
    Burstable

    Regional Pricing Comparison

    Compare a2-megagpu-16g (Linux) pricing across all regions - sorted by cost

    Current region (us-south1)Other regions
    💡 Click any bar to open that region in a new tab