(Tested) ComputeMark 2.1: GTX 480 vs HD 5870

ComputeMark 2.1 - GTX 480
ComputeMark 2.1 – GTX 480 – 1024×600 – normal mode

ComputeMark 2.1 - Julia 4D
ComputeMark 2.1 – Julia4D



The new version of ComputeMark is available. ComputeMark is a GPU computing benchmark based on Microsoft DirectX 11 DirectCompute technology. This new version brings new DirectCompute-based demos and 3 presets (normal, complex and extreme).

You can download the latest version here:
[download#132#image]

Okaaayyy, let’s do a quick test…

Testbed:
– CPU: Core i7 960 (default clocks)
– RAM: 4GB DDR3 corsair dominator 1600MHz
– Motherboard: GIGABYTE GA-X58A-UD5
– Win7 64-bit
– PSU: Antex TPQ 850W
– Display drivers: R257.15 (NVIDIA) and Catalyst 10.5 (ATI)

GeForce GTX 480

– 1024×600 windowed

  • Normal mode: 2932
  • Complex mode: 1693
  • Extreme mode: crash, crash and CRASH!

– 1920×1080 fullscreen

  • Normal mode: 1291
  • Extreme mode: re-crash…

I also tested with two GTX 480 in SLI but SLI mode is not supported (not in profiles)…

Radeon HD 5870

– 1024×600 windowed

  • Normal mode: 3229
  • Complex mode: 1855
  • Extreme mode: 307

– 1920×1080 fullscreen

  • Normal mode: 1549
  • Extreme mode: 238

ComputeMark 2.1 - HD 5870
HD 5870 – 1920×1080 – extreme mode

[via]

Related posts:

  1. (Tested) Unigine Heaven 2.1: GTX 480 vs GTX 470 vs HD 5870 in OpenGL 4.0 and Direct3D 11 in Extreme Tessellation
  2. ComputeMark: DX11 DirectCompute Fluid Simulation Benchmark
  3. (Tested) OpenGL 4.0 and Direct3D 11 Tessellation: GTX 480 vs GTX 470 vs HD 5870 vs HD 5770
  4. (Tested) AvP DX11 Tessellation Battle: GTX 480 vs GTX 470 vs HD 5870 vs HD 5770
  5. (Tested) 3DMARK11 DX11 Battle: GTX 580 vs GTX 480 vs HD 6870 vs HD 5870 vs GTX 460 vs HD 5770

10 thoughts on “(Tested) ComputeMark 2.1: GTX 480 vs HD 5870”

  5. DrBalthar

    People just have a hard time to grasp that GTX4xx series is an epic failure. And bringing their GPGPU marketing bollocks onto an equal fairground (meaning Open APIs not closed ones) it shows it even more obviously.

  7. zlat

    Great program Robert. I’m really happy to see two algorithm for the Fluid simulation.

    btw … anyone know why are the Fermi is so slow in 3d texturing?

  8. Robert Varga

    Zlat, they have said that they know that it’s a problem on Fermi, and that they “have to sort it out (3d texture performance)”.

    They didn’t wrote if they are speaking about Fermi or Fermi 2.

  9. Robert Varga

    Btw now they are working on SLI profile. Also I don’t know when it will be ready, but I will place some info on the computemark homepage when they will give me some new info.

  10. Pingback: AMD HD69xx leaks and news thread

Comments are closed.