NVIDIA H100 SXM5 vs NVIDIA RTX A5500. Specifications

Comparison of basic technical data of graphics cards NVIDIA H100 SXM5 and NVIDIA RTX A5500, chip, information processing units.

• Architecture
  Hopper Ampere
• Execution Units
  8448 80
• GPU Chip
  GH100 GA102
• L1 Cache
  192 KB 128 KB
• L2 Cache
  50 MB 6 MB
• Render Output Units
  24 96
• Texture Units
  528 320
• Shader
  - 10240

Video memory capacity, type, bus width and bandwidth describe different properties. These values alone do not determine graphics performance.

• Memory Bandwidth
  1,920 GB/s 768 GB/s
• Memory Bus
  5120 bit 384 bit
• Memory Clock
  1500 MHz 2000 MHz
• Memory Data Rate
  3 Gbps 16 Gbps
• Memory Size
  80 GB 24 GB
• Memory Type
  HBM3 GDDR6

Published GPU clock frequencies describe different operating modes. Clock speed alone does not determine graphics performance.

• Base Clock
  1065 MHz 1170 MHz
• Boost Clock
  1780 MHz 1695 MHz

Graphics card power limits and recommended system power supply describe different specifications.

• Board Number
  PG520 -
• System Power Supply
  1100 W (PSU recommendation) -
• TDP
  700 W 230 W

Connectivity and connections.

• Max. Displays
  1 -
• PCIe Interface
  PCIe 5.0 x16 PCIe 4.0 x16
• Power Connectors
  8-pin EPS 1x 8-pin
• DisplayPort Ports
  - 4 × DisplayPort 1.4a

Technical data that is used to its full potential in computer games.

• CUDA
  9.0 -
• DirectX
  No 12_2
• Max. resolution
  3840x2160@60Hz 7680 x 4320 at 60 Hz
• OpenCL
  3.0 3.0
• OpenGL
  No 4.6
• Shader Model
  No 6.6
• Vulkan
  No 1.3

The difference in size, weight and slot of the compared devices

• Length
  - 267 mm
• Slot Size
  - Dual-slot
• Width
  - 112 mm

Comparison of interfaces and release dates for GPU NVIDIA RTX A5500 and NVIDIA H100 SXM5.

• Process Technology
  4 nm 8 nm
• Release date
  Q2/2022 -

• FP16 (theoretical)
  120.3 TFLOPS 34.71 TFLOPS
• FP32 (theoretical)
  30.07 TFLOPS 34.71 TFLOPS
• FP64 (theoretical)
  15.04 TFLOPS 1,085 GFLOPS
• Pixel Rate
  42.72 GPixel/s 162.7 GPixel/s
• Texture Rate
  939.8 GTexel/s 542.4 GTexel/s