PNY NVIDIA Quadro P4000

I run Mastercam 2019 and this is the best card for the money imo. I haven’t seen loads to even warrant an upgrade from this and the parts i program can be pretty hardcore with hemstitching. So far there’s nothing that bogs it down. I paired it with 32gb ram and a core i9-9900k and it seems the processor is going to become the bottleneck if anything, which is saying a lot because that core i9 is rad.

I have owned the P4000 quadro since Feb 2018, bought from Amazon. I just installed it to a second motherboard, a Gigabyte X570 Master with the new PCIE 4.0 bandwidth. The prior motherboard was an Asus M5A99FX Pro 2.0. It is performing flawlessly powering three 32 inch 4K monitors from monoprice at max resolution. Beautiful, Beautiful, Beautiful! I use it mainly to run a trading platform called Think or Swim which has many small dots, squares, dashes, lines and numerous trading indicators, etc. The card works very well with windows 7 or windows 10 drivers but I just google ” Nvidia P4000 drivers” and download and install and you will have more options with the Nvidia drivers. I wish all installations would be as easy as this card.

I am a film editor. This graphics card smoothly renders 4k videos.

Item did not come in retail packaging and did not include all the item that were listed as included. Software Installation Disc, Printed Quick Start Guide, Stereo Connector Bracket were not in box. Card did look to be in new condition and works fine . This is not how I like to receive electronics.

After 4 years of talking myself out of buying a Quadro – I finally broke down and selected this beautiful balance of OpenGL brilliance. The PASCAL series is excellent technology – but, I was shocked at the speed of this card. (Very smooth transitions – video capture/editing; excellent OpenGL support in rendering and simulation software). I’m currently test-driving the newest Blender software.
I use Solid Edge / Works extensively (which I consider both a parametric-marvel) – and, I needed to get pass the ridiculous hang-ups that I was experiencing while working with large files with face sets during surface modeling (not always a memory issue!).
In Solid Edge, I try to avoid using the Sweep command; I’ll practice the ‘built-in’ face-set command and grab the face/profile of objects (which is excellent for piping systems and merging complex walls / boundaries). However, the GTX card’s limited OpenGL functions are obvious after about the 50th face-set down the tree. My advanced files used to require that I disable the ‘shaded with lines’ command – because the lines (vectorization) require extensive OpenGL resources and memory (which causes the detail to suffer a little; bad stuff for surface modeling)
Quadro is designed for accuracy, detail, and heavy vector calculation; GTX is designed for shooting things at 100 frames/second
OpenGL support is premium – and, the Quadro card will unlock drivers (OpenCL) for those looking to develop their own software (i.e. Blender, 3Ds Max)
INTEL 6700k (liquid cooled)
ASUS Z170 Sabertooth
Samsung Pro 950 (with Windows 10 Pro); make sure you have the latest Samsung NVMe drivers installed (Windows will install generic)
16GB DDR4 – 3000 MHz – Crucial Ballistix Elite memory (I used this stable and durable profile in all my systems)
Nothing else really matters – this is a very efficient system with a Quadro Pascal card

***If you switch from GTX to Quadro – make sure you grab the Quadro drivers and make the appropriate settings in your software (OpenGL will need to be set manually in NVidia control panel) – beautiful things will unfold. Also, try to match your CPU and Memory speed to a synchronous value (i.e. 2133cpu & 2133memory) – ***UPDATE: synchronous is a reference to stability and accuracy of data – not normally associated with overclocking (in professional use)

Excelente precio, el producto llego en tiempo y forma, hasta ahora va desquitando la inversión, muy buen producto los modelos 3D grandes los mueve muy fluidos, lo que si es que la tarjeta supera los 90ºC cuando esta siendo forzada.

This is as good of performance as can be had in a single-wide card. The card is pretty thin and nott too terribly long. The fan is very quiet. The card is highly configurable through the NVSMI. The PLL is very solid, locking fast and consistently to a myriad of clock frequencies. The performance for CUBLAS (CUDA Basic Linear Algebra System) is very impressive. While it is still not out performing my ancient K20 Tesla, it is close, and it is considerably smaller and quieter and considerably more efficient and versatile (can be used as a display card, a “Tesla compute ‘cluster'”* or both. *(A confusing term NVidia uses for the operational mode — TCC dedicates the card entirely to compute.)