NVIDIA A40 vs NVIDIA Tesla T4

Comparative analysis of NVIDIA A40 and NVIDIA Tesla T4 videocards for all known characteristics in the following categories: Essentials, Technical info, Video outputs and ports, Compatibility, dimensions and requirements, API support, Memory. Benchmark videocards performance analysis: Geekbench - OpenCL, PassMark - G2D Mark, PassMark - G3D Mark, CompuBench 1.5 Desktop - Face Detection (mPixels/s), CompuBench 1.5 Desktop - Ocean Surface Simulation (Frames/s), CompuBench 1.5 Desktop - T-Rex (Frames/s), CompuBench 1.5 Desktop - Video Composition (Frames/s), CompuBench 1.5 Desktop - Bitcoin Mining (mHash/s), GFXBench 4.0 - Car Chase Offscreen (Frames), GFXBench 4.0 - Manhattan (Frames), GFXBench 4.0 - T-Rex (Frames), GFXBench 4.0 - Car Chase Offscreen (Fps), GFXBench 4.0 - Manhattan (Fps), GFXBench 4.0 - T-Rex (Fps).

 

Differences

Reasons to consider the NVIDIA A40

  • Videocard is newer: launch date 2 year(s) 0 month(s) later
  • Around 30% higher core clock speed: 1305 MHz vs 1005 MHz
  • Around 15% higher boost clock speed: 1740 MHz vs 1515 MHz
  • A newer manufacturing process allows for a more powerful, yet cooler running videocard: 8 nm vs 12 nm
  • 3.2x better performance in Geekbench - OpenCL: 193429 vs 61276
  • Around 6% better performance in PassMark - G2D Mark: 627 vs 590
  • Around 35% better performance in PassMark - G3D Mark: 14665 vs 10833
Specifications (specs)
Launch date 5 Oct 2020 vs 13 September 2018
Core clock speed 1305 MHz vs 1005 MHz
Boost clock speed 1740 MHz vs 1515 MHz
Manufacturing process technology 8 nm vs 12 nm
Benchmarks
Geekbench - OpenCL 193429 vs 61276
PassMark - G2D Mark 627 vs 590
PassMark - G3D Mark 14665 vs 10833

Reasons to consider the NVIDIA Tesla T4

  • 4x lower typical power consumption: 75 Watt vs 300 Watt
  • 5.5x more memory clock speed: 10000 MHz vs 1812 MHz (14.5 Gbps effective)
Thermal Design Power (TDP) 75 Watt vs 300 Watt
Memory clock speed 10000 MHz vs 1812 MHz (14.5 Gbps effective)

Compare benchmarks

GPU 1: NVIDIA A40
GPU 2: NVIDIA Tesla T4

Geekbench - OpenCL
GPU 1
GPU 2
193429
61276
PassMark - G2D Mark
GPU 1
GPU 2
627
590
PassMark - G3D Mark
GPU 1
GPU 2
14665
10833
Name NVIDIA A40 NVIDIA Tesla T4
Geekbench - OpenCL 193429 61276
PassMark - G2D Mark 627 590
PassMark - G3D Mark 14665 10833
CompuBench 1.5 Desktop - Face Detection (mPixels/s) 127.622
CompuBench 1.5 Desktop - Ocean Surface Simulation (Frames/s) 3026.812
CompuBench 1.5 Desktop - T-Rex (Frames/s) 18.798
CompuBench 1.5 Desktop - Video Composition (Frames/s) 119.936
CompuBench 1.5 Desktop - Bitcoin Mining (mHash/s) 910.721
GFXBench 4.0 - Car Chase Offscreen (Frames) 14076
GFXBench 4.0 - Manhattan (Frames) 1976
GFXBench 4.0 - T-Rex (Frames) 1781
GFXBench 4.0 - Car Chase Offscreen (Fps) 14076
GFXBench 4.0 - Manhattan (Fps) 1976
GFXBench 4.0 - T-Rex (Fps) 1781

Compare specifications (specs)

NVIDIA A40 NVIDIA Tesla T4

Essentials

Architecture Ampere Turing
Code name GA102 TU104
Launch date 5 Oct 2020 13 September 2018
Place in performance rating 55 279
Type Workstation

Technical info

Boost clock speed 1740 MHz 1515 MHz
Core clock speed 1305 MHz 1005 MHz
Manufacturing process technology 8 nm 12 nm
Peak Double Precision (FP64) Performance 1169 GFLOPS (1:32)
Peak Half Precision (FP16) Performance 37.42 TFLOPS (1:1)
Peak Single Precision (FP32) Performance 37.42 TFLOPS
Pipelines 10752
Pixel fill rate 194.9 GPixel/s
Texture fill rate 584.6 GTexel/s
Thermal Design Power (TDP) 300 Watt 75 Watt
Transistor count 28300 million 13,600 million

Video outputs and ports

Display Connectors 3x DisplayPort No outputs

Compatibility, dimensions and requirements

Form factor Dual-slot
Interface PCIe 4.0 x16 PCIe 3.0 x16
Length 267 mm (10.5 inches)
Recommended system power (PSU) 700 Watt
Supplementary power connectors 8-pin EPS None
Width 112 mm (4.4 inches)

API support

DirectX 12.2 12.0 (12_1)
OpenCL 3.0
OpenGL 4.6 4.6
Shader Model 6.6
Vulkan

Memory

Maximum RAM amount 48 GB
Memory bandwidth 695.8 GB/s
Memory bus width 384 bit
Memory clock speed 1812 MHz (14.5 Gbps effective) 10000 MHz
Memory type GDDR6