Where possible, I've run 3DMark benchmarks on old and new graphics cards... this can be useful to see how much difference there is between your old card and new card, when upgrading a graphics card, or also when upgrading to a newer system. Scroll down and you'll find vintage graphics card scores as well.
3DMark 2006 (DX9)
Latest graphics card: GTX 1060 3GB: 23,346*
Previous graphics card: GTX 460: 22,876*
3DMark 2011 (DX11)
Latest graphics card: GTX 1060 3GB: P 11,821*
Previous graphics card: GTX 460: P 3635*
* System: i5 4590, 3.3-3.7GHz, 8gb DDR3, Windows 10
** System: Core2Duo E4600, 2.4GHz, 4GB DDR2, Windows 7, 'Win10
*** System: Pentium M 1.8GHz, 1GB (DDR), AGP
x = May not have been optimal drivers
^ Passive
% increase is increase in performance over system listed one place below, based on figures in bold or figures combined as an average
3DMark Versions, and the Direct X version / graphics cards they're designed to test:
Tested with a Windows 98SE, AMD K6-2 500MHz, 256mb SDRAM system. *CPU, non 3D accelerated Quake 640x480 resolution. ** OpenGL driver followed by 3dfx Glide driver for Quake, same 640x480 resolution.
3DMark 2006 (DX9)
Latest graphics card: GTX 1060 3GB: 23,346*
Previous graphics card: GTX 460: 22,876*
3DMark 2011 (DX11)
Latest graphics card: GTX 1060 3GB: P 11,821*
Previous graphics card: GTX 460: P 3635*
3DMark Version: | TimeSpy v1.0 | FireStrike | 2011 | 2006 | 2005 | 2003 | 2001SE |
Geforce GTX 1060 3GB* +225% | 3,762 | 9,607 | 11,821 | 23,348 | 21,998 | 124,611 | 63,025 |
Geforce GTX 460 2GB* +395% | - | - | 3,635 | 22,876 | - | - | - |
Geforce GTX 460 2GB** +71% ^ Windows 7 / ' Windows 10 | - | - | - | 4,614 | 14,634 11,785' | 48,378 41,933' | 25,940 13,709' |
Radeon X1900 XT 512MB** | - | - | - | 5,123 | 10,255 | 16,762 | 22,564 |
Radeon HD5450^ 1GB** ('W10) | - | - | - | 2,617 | 4,756 | 7,576 | - |
Geforce 7300GT 256MB** | - | - | - | 1,646 | 3,084 | 7,297 | 15,877 |
Radeon 9700 128MB*** | - | - | - | - | 1,695 | 4,564 | 14,218 |
Radeon X300XE 128MB** | - | - | - | - | 942(x) | 1,808(x) | 5,501(x) |
Intel G965 (Built-in to motherboard)** | - | - | - | - | 884(x) | 1,659(x) | 4,246(x) |
* System: i5 4590, 3.3-3.7GHz, 8gb DDR3, Windows 10
** System: Core2Duo E4600, 2.4GHz, 4GB DDR2, Windows 7, 'Win10
*** System: Pentium M 1.8GHz, 1GB (DDR), AGP
x = May not have been optimal drivers
^ Passive
% increase is increase in performance over system listed one place below, based on figures in bold or figures combined as an average
3DMark Versions, and the Direct X version / graphics cards they're designed to test:
- 3DMark 2000: DirectX7
- 3DMark 2001 SE: DirectX8.1
- 3DMark 2003: DirectX9.0a
- 3DMark 2005: DirectX9.0c - Last version that you could view results without the internet (v120, not v130 - v130 requires internet).
- 3DMark 2006: DirectX9.0c - Needs an internet connection to view results
- 3DMark Vantage: DirectX10
- 3DMark2011: DirectX11
- 3DMark TimeSpy 1.0: DirectX12
- 3DMark FireStrike: DirectX11
"Vintage" Graphics Cards
Travelling further back in time, here are some of the results for 3DMark 2000 and 99Max, and some much older vintage graphics cards:3DMark (or other) | 2001 | 2000 | 99Max | QuakeGL (Timedemo fps) |
3dfx Voodoo3 3000 16mb AGP | 528 | 1343 | 2484 | 86.3fps (**) 101.1fps (3dfx**) |
GeForce 4 MX420 64mb AGP | .. | 1421 | 2481 | 71fps (VGA/16) |
GeForce 2 MX/MX400 64mb AGP | .. | 1391 | 2514 | 70fps (VGA/16) |
NVidia FX5200 128mb AGP | .. | 1438 | 2295 | 68.5fps (VGA/16), 23.8fps* |
NVidia Vanta / Vanta LT 16mb AGP | .. | 1124 | 2547 | 39.9fps (VGA/16) |
.. | ... | - | - | -fps (-) |
Tested with a Windows 98SE, AMD K6-2 500MHz, 256mb SDRAM system. *CPU, non 3D accelerated Quake 640x480 resolution. ** OpenGL driver followed by 3dfx Glide driver for Quake, same 640x480 resolution.
Comments