Skip to main content

L4 · Chips

Data-Center GPU Comparison

The merchant data-center accelerators behind AI training and inference — NVIDIA Hopper/Blackwell vs AMD CDNA — compared on memory, bandwidth, and power. The interesting deltas are increasingly about HBM capacity and bandwidth, not raw FLOPs.

VendorAcceleratorArchitectureMemoryBandwidthTDPYear
NVIDIAH100 (SXM)Hopper80 GB HBM33.35 TB/s700 W2022
NVIDIAH200 (SXM)Hopper141 GB HBM3e4.8 TB/s700 W2024
NVIDIAB200Blackwell192 GB HBM3e8 TB/s1000 W2024–25
NVIDIAGB200 (superchip)Grace + 2× Blackwell384 GB HBM3e16 TB/s~2700 W2025
AMDMI300XCDNA 3192 GB HBM35.3 TB/s750 W2023
AMDMI325XCDNA 3256 GB HBM3e6.0 TB/s1000 W2024–25
AMDMI355XCDNA 4288 GB HBM3e8 TB/s~1400 W2025

Specs consolidated from NVIDIA and AMD datasheets. Per-SKU figures vary by form factor (SXM / OAM / PCIe).

The merchant GPU is only half the story

Hyperscaler custom ASICs now grow faster than merchant GPUs by unit volume, and HBM + advanced packaging gate how many of any of these actually ship. The sourced landscapes:

Engineering reference, not investment advice.