NVIDIA Quadro FX 3600M vs NVIDIA Quadro FX 370M vs NVIDIA Quadro FX 3500M
NVIDIA Quadro FX 3600M► remove
The NVIDIA Quadro FX 3600M is a professional graphics card for CAD, DCC and other modeling applications based upon the 8800M GTX consumer card (G92 architecture). It was a high-end card at the time of announcement. Certified drivers grant error-free and faster performance in professional applications. The FX 3600M was later succeeded by the more efficient 55nm FX 2800M and FX 3800M, both of which still utilize the G92 architecture.
The most important aspect of the Quadro FX 3600M is its use "Unified Shaders". There no longer exist any dedicated pixel or vertex shaders. Instead, the FX 3600M uses 64 so-called stream processors to handle most of the graphics work previously processed by pixel and vertex shaders. Furthermore, the shader units are clocked higher than the GPU core (1250 MHz).
The performance of the FX 3600M shoul be higher than the 1600M due to the 256 bit wide memory bus and the high number of Unified Shaders. According to Nvidia, the GPU is best suited for modelling large 3D scenes and complex models.
NVIDIA Quadro FX 370M► remove
The NVIDIA Quadro FX 370M is an entry level graphics card for mobile workstations (laptops). It is based on the same core as the Quadro NVS 160M and the GeForce 9200M / 9300M GS. The drivers are optimized for OpenGL applications and professional 3D applications like CAD, DCC, and visualization applications. Furthermore, the drivers are certified to work with the most popular apps without problems.
The gaming performance is a bit below the similar 9200M / 9300M chip because of the professional drivers and therefore the GPU is only capable of displaying low end games fluently. Modern intensive 3D games wont run fluently or only with very low detail settings.
As the 9300M series, the Quadro FX 370M supports PureVideo HD to decode (HD) videos in the GPU.
The performance of the chip also depends on the used graphics memory. Either fast GDDR3 memory (clocked at 700 MHz) or slow GDDR2 (with 400MHz) can be used by the laptop manufacturer.
In the Dell Precision M2400, only GDDR2 graphics memory is used, leading to a slow performance (e.g. Doom3 - ultra 40fps). According to Dell the graphics card needs 20 Watt.
NVIDIA Quadro FX 3500M► remove
The NVIDIA Quadro FX 3500M was the top professional graphic solution for the mobile use in 2007. The Quadro FX 3500M is especially suitable for CAD, digital image-, video editing and scientifical visualizations (according NVIDIA). It is based upon the NVIDIA GeForce Go 7950GTX. Therefore also a good gaming performance can be expected (for 2007 and if modded drivers are used). The card is shipped with certified drivers for professional applications, which are not optimized for games (inofficially the manual installation of Forceware drivers should be possible).
The drivers of the Quadro FX series are certified for CAD applications (e.g. AutoCad, Inventor, Solidstation, Solud Edge, ...) and DCC applictions (e.g. Maya, 3DS Max, Lightwave 3D, ...) and offers a 12-Bit precision for Subpixel, full image Antialiasing, hardware accelerated Antialiasing for dots and lines and additional optimized settings, which are optimized for CAD and DCC.
Nvidia describes the 3000 series as a power package for big laptops and data structures.
|NVIDIA Quadro FX 3600M||NVIDIA Quadro FX 370M||NVIDIA Quadro FX 3500M|
|Quadro FX Series|
|96 -||8 -||24 / 8|
|Core||500 MHz||580 MHz||575 MHz|
|Shader||1250 MHz||1450 MHz|
|Memory||800 MHz||700 MHz||700 MHz|
|Bus||256 Bit||64 Bit||256 Bit|
|Max. Memory||512 MB||256 MB||512 MB|
|DirectX||DirectX 10, 4.0||DirectX 10, 4.0||DirectX 9c, 3.0|
|Transistors||754 Million||278 Million|
|Technology||65 nm||65 nm||90 nm|
|Features||PureVideo technology (H.264, VC-1, MPEG2, WMV9 decoding acceleration), HDCP-capable, PowerMizer 7.0 energy administration (dynamic switching between performance and energy saving), HDR (High Dynamic-Range Lighting), designed for Windows Vista, supports Dual-Link DVI-D exits for TFT resolutions up to 2560x1600, LVDS, VGA, HDTV, HDCP, HDMI, PCI-E 16x, OpenGL 2.1, Gigathread technology||12-bit-sub-pixel, 32-bit-floating-point precision, 32-bit floating point precision, 128 bit graphic pipeline, hardware 3D window clipping, high precision dynamic-range (DPDR) technology, unified memory architecture, nView multi-display (for multi-monitor support), PowerMizer options for energy saving, architecture which is optimized for workstations, rotated-grid full-scene antialiasing (RG FSAA), unified driver architecture, NVIDIA PureVideo technology|
|MXM Factor III||OpenGL 2.1, max. 1920x1200 (digital) bzw 2048x1536 (analog), max. 2 unabhängige Displays, PCI-Express x16, optionale Interfaces (des Chips): DP, VGA, DVI, LVDS, HDMI|