NVIDIA H100 PCIe vs NVIDIA GeForce RTX 3090. Specifications, performance, tests

Comparison of basic technical data of graphics cards NVIDIA H100 PCIe and NVIDIA GeForce RTX 3090, chip, information processing units.

• GPU Chip
  GH100 GA102-300-A1
• Execution units
  7296 82
• Render Output Units
  24 112
• Texture Units
  456 328

Comparison of the amount of memory on board graphics cards. The more the better.

• Memory Size
  80 GB 24 GB
• Memory Type
  HBM2e GDDR6X
• Memory Clock
  1 GHz 1.219 GHz
• Memory Speed
  2 Gbps 19.5 Gbps
• Memory Bandwith
  1,280 GB/s 936 GB/s
• Memory Interface
  5120 bit 384 bit

Let's compare the memory frequency of graphics cards NVIDIA H100 PCIe and NVIDIA GeForce RTX 3090. The higher the better.

• Base Clock
  1.065 GHz 1.4 GHz
• Boost Clock
  1.65 GHz 1.695 GHz

Connectors, the number of thermal watts emitted in normal mode and at overclocking.

• TDP
  350 W 350 W

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

• Max. Displays
  1 4
• DirectX
  No 12_2
• OpenGL
  No 4.6
• OpenCL
  3.0 3.0
• Vulkan
  No 1.3
• Shader Model
  No 6.6

Comparison of interfaces and release dates for GPU NVIDIA GeForce RTX 3090 and NVIDIA H100 PCIe.

• Architecture
  Hopper Ampere
• GPU Interface
  PCIe 5.0 x16 PCIe 4.0 x 16
• Structure size
  4 nm 8 nm
• Release date
  Q2/2022 Q3/2020

• Pixel Rate
  39.60 GPixel/s 189.8 GPixel/s
• Texture Rate
  752.4 GTexel/s 556.0 GTexel/s
• FP16 (half)
  96.31 TFLOPS 35.58 TFLOPS
• FP32 (float)
  24.08 TFLOPS FP32 (float)
• FP64 (double)
  12.04 TFLOPS FP64 (double)