First Look: Nvidia Quadro M2000 GPU - Nvidia has CAD users firmly in its sights with new release

Published 19 April 2016

Posted by greg corke

Article tagged with: solidworks, gpu, workstations, graphics cards, nvidia quadro, iray

The Nvidia Quadro M2000 - sitting pretty on our desk, itching to get put through its paces in our suite of 3D CAD software tests.


*****************************************************************************************

Update

Check out our full review of the Nvidia Quadro M2000 with performance testing for SolidWorks, PTC Creo, Autodesk Fusion 360 and SolidWorks Visualize.

*****************************************************************************************


Today, Nvidia added a new graphics card to its family of professional GPUs. The 4GB Quadro M2000 should hit the sweet spot for 3D CAD, with applications like SolidWorks, PTC Creo and Siemens NX, and will replace the company’s current mid-range offering, the Quadro K2200.

In Nvidia nomenclature the ‘M’ stands for Maxwell, the codename for the company’s current GPU microarchitecture, which replaced Kepler (K) in late 2014.

The Quadro M2000 fleshes out the ‘official’ Maxwell Quadro line-up, joining the higher end Quadro M4000, M5000, M6000, which were all introduced last year.

We say ‘official’, because even though the Quadro K2200 was ‘Kepler’ in name, it was actually built from Maxwell technology, and was part of a mixed architecture rollout of K Series GPUs designed to coincide with the launch of Intel’s ‘Haswell-based’ Xeon E5-series CPUs and new workstations from the major OEMs.

So what can CAD users expect from the new Quadro M2000? While Nvidia’s published SPECapc for SolidWorks 2015 graphics composite benchmark figures only show a 7% to 10% performance increase over the Quadro K2200, we expect there to be a more significant increase in workflows where the GPU is pushed more, and less constrained by the speed of the CPU.

Indeed, with a quoted peak single precision compute performance of up to 1.8 TFLOPS, a 38% increase over the Nvidia Quadro K2200 (1.3 TFLOPS), the Quadro M2000 should start to show its true potential when ray trace rendering with Nvidia Iray.

With Nvidia currently making a big push in the CAD market with its physically based Iray rendering technology, through applications like SolidWorks Visualize, and integrations for Siemens NX, Rhino, Cinema 4D and 3ds Max, having the additional compute performance on tap should make a significant dent in render times.

In terms of specs the Quadro M2000 has 768 CUDA cores and up to 106GB/s memory bandwidth (compared to the Quadro K2200’s 640 CUDA cores and 80GB/s). It also has a maximum power consumption of 75W, a step up from the K2200’s 60W, and uses the PCIe 3.0 graphics bus, rather than PCIe 2.0.

It boasts four DisplayPort 1.2 connectors, enabling it to support up to four 4K monitors @ 60Hz. Additionally, the Quadro M2000 can support up to two 5K monitors @60Hz, or a single 8K projector.

The Quadro M2000 should be offered by all the major workstation OEMs this month. We already have a sample card in our possession, and will be testing it out over the coming weeks, so check back soon for our verdict.

Comments:

next time

Posted by KyleLE on 22 April 2018 at 01:28 AM

Leave a comment

Enter the word you see below: