AMD Radeon Pro Vega 48 vs NVIDIA Tesla P40

Comparative analysis of AMD Radeon Pro Vega 48 and NVIDIA Tesla P40 videocards for all known characteristics in the following categories: Essentials, Technical info, Video outputs and ports, Compatibility, dimensions and requirements, API support, Memory, Technologies. Benchmark videocards performance analysis: PassMark - G3D Mark, PassMark - G2D Mark, Geekbench - OpenCL, 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 - 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), CompuBench 1.5 Desktop - Video Composition (Frames/s).

 

Differences

Reasons to consider the AMD Radeon Pro Vega 48

  • Videocard is newer: launch date 2 year(s) 6 month(s) later
  • 679.4x more texture fill rate: 249.6 GTexel/s vs 367.4 GTexel / s
  • A newer manufacturing process allows for a more powerful, yet cooler running videocard: 14 nm vs 16 nm
  • Around 74% better performance in PassMark - G2D Mark: 784 vs 450
  • Around 25% better performance in GFXBench 4.0 - Car Chase Offscreen (Frames): 11925 vs 9538
  • Around 7% better performance in GFXBench 4.0 - Manhattan (Frames): 3947 vs 3699
  • 3.1x better performance in GFXBench 4.0 - T-Rex (Frames): 10411 vs 3340
  • Around 25% better performance in GFXBench 4.0 - Car Chase Offscreen (Fps): 11925 vs 9538
  • Around 7% better performance in GFXBench 4.0 - Manhattan (Fps): 3947 vs 3699
  • 3.1x better performance in GFXBench 4.0 - T-Rex (Fps): 10411 vs 3340
Specifications (specs)
Launch date 19 March 2019 vs 13 September 2016
Texture fill rate 249.6 GTexel/s vs 367.4 GTexel / s
Manufacturing process technology 14 nm vs 16 nm
Benchmarks
PassMark - G2D Mark 784 vs 450
GFXBench 4.0 - Car Chase Offscreen (Frames) 11925 vs 9538
GFXBench 4.0 - Manhattan (Frames) 3947 vs 3699
GFXBench 4.0 - T-Rex (Frames) 10411 vs 3340
GFXBench 4.0 - Car Chase Offscreen (Fps) 11925 vs 9538
GFXBench 4.0 - Manhattan (Fps) 3947 vs 3699
GFXBench 4.0 - T-Rex (Fps) 10411 vs 3340

Reasons to consider the NVIDIA Tesla P40

  • Around 9% higher core clock speed: 1303 MHz vs 1200 MHz
  • Around 18% higher boost clock speed: 1531 MHz vs 1300 MHz
  • Around 25% higher pipelines: 3840 vs 3072
  • 50x lower typical power consumption: 250 Watt vs 12500 million
  • 3x more maximum memory size: 24 GB vs 8 GB
  • 12.7x more memory clock speed: 10008 MHz vs 786 MHz (1572 MHz effective)
  • Around 12% better performance in PassMark - G3D Mark: 12678 vs 11300
  • 2.2x better performance in CompuBench 1.5 Desktop - Face Detection (mPixels/s): 300.355 vs 135.416
  • Around 16% better performance in CompuBench 1.5 Desktop - Ocean Surface Simulation (Frames/s): 3559.383 vs 3063.269
  • Around 56% better performance in CompuBench 1.5 Desktop - T-Rex (Frames/s): 19.757 vs 12.678
  • Around 37% better performance in CompuBench 1.5 Desktop - Bitcoin Mining (mHash/s): 1214.142 vs 883.876
Specifications (specs)
Core clock speed 1303 MHz vs 1200 MHz
Boost clock speed 1531 MHz vs 1300 MHz
Pipelines 3840 vs 3072
Thermal Design Power (TDP) 250 Watt vs 12500 million
Maximum memory size 24 GB vs 8 GB
Memory clock speed 10008 MHz vs 786 MHz (1572 MHz effective)
Benchmarks
PassMark - G3D Mark 12678 vs 11300
Geekbench - OpenCL 53461 vs 53429
CompuBench 1.5 Desktop - Face Detection (mPixels/s) 300.355 vs 135.416
CompuBench 1.5 Desktop - Ocean Surface Simulation (Frames/s) 3559.383 vs 3063.269
CompuBench 1.5 Desktop - T-Rex (Frames/s) 19.757 vs 12.678
CompuBench 1.5 Desktop - Bitcoin Mining (mHash/s) 1214.142 vs 883.876

Compare benchmarks

GPU 1: AMD Radeon Pro Vega 48
GPU 2: NVIDIA Tesla P40

PassMark - G3D Mark
GPU 1
GPU 2
11300
12678
PassMark - G2D Mark
GPU 1
GPU 2
784
450
Geekbench - OpenCL
GPU 1
GPU 2
53429
53461
CompuBench 1.5 Desktop - Face Detection (mPixels/s)
GPU 1
GPU 2
135.416
300.355
CompuBench 1.5 Desktop - Ocean Surface Simulation (Frames/s)
GPU 1
GPU 2
3063.269
3559.383
CompuBench 1.5 Desktop - T-Rex (Frames/s)
GPU 1
GPU 2
12.678
19.757
CompuBench 1.5 Desktop - Bitcoin Mining (mHash/s)
GPU 1
GPU 2
883.876
1214.142
GFXBench 4.0 - Car Chase Offscreen (Frames)
GPU 1
GPU 2
11925
9538
GFXBench 4.0 - Manhattan (Frames)
GPU 1
GPU 2
3947
3699
GFXBench 4.0 - T-Rex (Frames)
GPU 1
GPU 2
10411
3340
GFXBench 4.0 - Car Chase Offscreen (Fps)
GPU 1
GPU 2
11925
9538
GFXBench 4.0 - Manhattan (Fps)
GPU 1
GPU 2
3947
3699
GFXBench 4.0 - T-Rex (Fps)
GPU 1
GPU 2
10411
3340
Name AMD Radeon Pro Vega 48 NVIDIA Tesla P40
PassMark - G3D Mark 11300 12678
PassMark - G2D Mark 784 450
Geekbench - OpenCL 53429 53461
CompuBench 1.5 Desktop - Face Detection (mPixels/s) 135.416 300.355
CompuBench 1.5 Desktop - Ocean Surface Simulation (Frames/s) 3063.269 3559.383
CompuBench 1.5 Desktop - T-Rex (Frames/s) 12.678 19.757
CompuBench 1.5 Desktop - Bitcoin Mining (mHash/s) 883.876 1214.142
GFXBench 4.0 - Car Chase Offscreen (Frames) 11925 9538
GFXBench 4.0 - Manhattan (Frames) 3947 3699
GFXBench 4.0 - T-Rex (Frames) 10411 3340
GFXBench 4.0 - Car Chase Offscreen (Fps) 11925 9538
GFXBench 4.0 - Manhattan (Fps) 3947 3699
GFXBench 4.0 - T-Rex (Fps) 10411 3340
CompuBench 1.5 Desktop - Video Composition (Frames/s) 204.32

Compare specifications (specs)

AMD Radeon Pro Vega 48 NVIDIA Tesla P40

Essentials

Architecture Vega 10 PRO Pascal
Code name Greenland GP102
GCN generation GCN 5.0
Launch date 19 March 2019 13 September 2016
Place in performance rating 228 229
Type Desktop Workstation
Launch price (MSRP) $5,699

Technical info

Boost clock speed 1300 MHz 1531 MHz
Core clock speed 1200 MHz 1303 MHz
Manufacturing process technology 14 nm 16 nm
Peak Double Precision (FP64) Performance 499.2 GFLOPS
Peak Half Precision (FP16) Performance 15.97 TFLOPS
Peak Single Precision (FP32) Performance 7.987 TFLOPS
Pipelines 3072 3840
Pixel fill rate 83.20 GPixel/s
Texture fill rate 249.6 GTexel/s 367.4 GTexel / s
Thermal Design Power (TDP) 12500 million 250 Watt
Floating-point performance 11,758 gflops
Transistor count 11,800 million

Video outputs and ports

Display Connectors No outputs No outputs

Compatibility, dimensions and requirements

Interface PCIe 3.0 x16 PCIe 3.0 x16
Length 267 mm
Supplementary power connectors 1x 6-pin + 1x 8-pin

API support

DirectX 12.1 12.0 (12_1)
OpenCL 2.0
OpenGL 4.6 4.6
Shader Model 6.3
Vulkan

Memory

High bandwidth memory (HBM)
Maximum RAM amount 8 GB 24 GB
Memory bandwidth 402.4 GB/s 480.4 GB / s
Memory bus width 2048 bit 384 Bit
Memory clock speed 786 MHz (1572 MHz effective) 10008 MHz
Memory type HBM2 GDDR5X

Technologies

Unified Video Decoder (UVD)
Video Code Engine (VCE)