amazon NVIDIA Quadro P5000 reviews
Overview
The launch of Pascal NVIDIA architecture, in addition to improving game performance, professional areas such as architecture, manufacturing, and design can greatly benefit from accelerating by GPUs and not unexpectedly, Pascal architecture has been equipped on the NVIDIA Quadro professional graphics card line.
Parameter comparison between Quadro P2000, Quadro P4000, Quadro P5000, Radeon Pro Duo
Based on GP106 GPU, Quadro P2000 is relatively close to GTX 1060. Comparing Shaders P2000 with GTX 1060 3GB, we find that it is less than 128 units and a slightly lower clock speed, but yes more memory. Similarly, the Quadro P4000 can be compared to the GTX 1070 less than 128 units and lower clock speed.
However, the Quadro P5000 has the same synchronization speed and clock speed as the 1080 GTX, but with twice the memory capacity, typical of professional graphics cards often work with larger data.
Quadro P5000 in a familiar shape. Perhaps you will notice the similarity between the design of the Quadro P5000 and GTX 1080. Because they have the same GPU, this is not surprising. However, video output is tweaked from GPU consumers. Here you will see 4 DisplayPort connectors and a Dual-Link DVI connector.
General computing performance – SPECviewperf12
SPECviewperf 12 is a benchmarking application centered around workstation graphics performance. Both OpenGL and DirectX features are measured by different workloads or “scenes”. It is important to note that the AMD Radeon Pro Duo seems to be running in single GPU mode for these tests. SPECviewperf does not provide any control for additional GPU targeting.
CATIA Viewset (Catia-04)
Catia-04 viewset is created from the simulation of the graphical workload created by CATIA V6 R2012 application from Dassault Systemes. The model size ranges from 5.1 to 21 million peaks.
This view includes multiple display modes supported by the application, including: wireframe, anti-aliasing, shaded with edges, depth of field, and ambient occlusion.
With SPECviewperf 12’s CATA view, we see a clear division between the different products in the Quadro product line. The Radeon Pro Duo runs in Single GPU mode struggling to compete with the low-end Quadro P2000.
Energy Viewset (Energy -01)
Energy-01 viewset is representative of a typical volume rendering application in seismic and oil fields. Similar to medical images such as MRI or CT, geophysical surveys create image slices through underground surfaces built in 3D grids. Display volume provides a 2D projection of 3D volume grid for further analysis and explanation.
At each frame, depending on the location of the viewer, a series of flat plates corresponding to the viewing angle are calculated on the CPU and then sent to the graphics hardware to calculate and calculate more like search function. The ability to convert, light and cut to reveal the internal structure. Finally, the slices are mixed together before the image is displayed.
In the energy workload we see a similar story. Performance among the Quadro options scales as you would expect, while the Radeon Pro Duo comes to the end despite having the powerful Fiji XT GPU.
Medical Viewset (Medical-01)
The medical-01 is representative of a typical mass display application for 2D projection of 3D volume grid. A typical 3D grid in this view is a group of 3D slices obtained by the scanner (such as CT or MRI).
In each frame, depending on the location of the viewer, a series of coplanar copies are aligned with the viewing angle calculated on the CPU and then sent to the graphics hardware for further calculations and calculations, such as Look up switching, lighting and cutting functions to display internal structures. Finally, the slices are mixed together before the image is displayed.
For the Medical viewset application, the performance is balanced with Quadro GPUs while the AMD card ends last.
Showcase Viewset (showcase-01)
Showcase-01 was created from the simulation of Autodesk’s Showcase 2013. The model used in the view includes 8 million vertices.
View mode is the first view in SPECviewperf for DX rendering. The display modes are included in the view including shading, projected shadows, and self-shadows.
By using standard DirectX for rendering, we see that the Showcase view is the first test to break Quadro’s superiority so far. Radeon Pro Duo in single GPU mode and very competitive with P4000 in this test.
Siemens NX (snx-02)
Snx-02 viewet was created from the simulation of the graphical workload generated by NX 8.0 applications from Siemens PLM. The model size ranges from 7.15 to 8.45 million peaks.
In testing with the Siemens NX view, Radeon Pro Duo achieved very low scores compared to Quadro rivals. This may indicate some issues with this workload and AMD’s GPU in contrast to Radeon Pro Duo’s relatively realistic performance levels.
Maya viewset (maya-04)
Maya viewset (maya-04) was created from simulating the graphical workload created by Maya 2013 application from Autodesk. Model size is 727,500 peaks.
The view includes multiple display modes supported by the application, including: shaded mode, ambient occlusion, multi-sample anti aliasing, and transparency.
In Maya viewset, we see that the Quadro P4000 and P5000 achieve similar results despite the performance gap of nearly 4 TFLOPs. This test uses Maya 2013, so it may not be optimized for high-capacity GPUs today.
3DS Max Viewset (3dsmax-05)
Models of this view come from SPECapc for 3ds Max 2015 standard and other sources. To make the best use of real-world use cases, some tests combine multiple viewing modes on the screen, each using a different display method. The display styles in the view reflect the most commonly used images in large markets, including reality, ghosting and wire frames. Some rendering modes are less used but interesting, such as faces, graphite and clay are also incorporated. The animation in the view is a combination of the spin model and the fly-through camera, depending on the model.
In 3ds Max using Nitrous DX11 renderer, Radeon Pro Duo in single GPU mode only achieves half of the points we see from Quadro P5000 despite the similarities in raw GPU calculations of these two products.
Solidworks viewset (sw-03)
The sw-03 viewset is generated from the SolidWorks 2013 SP1 simulation of Dassault Systemes. Models used in the context range from 2.1 to 21 million vertices.
The view includes multiple display modes supported by the application, including shaded mode, shaded with edges, obscured surroundings, slide show and environment map.
Performance Rendering – OpenCL and CUDA
LuxMark 3.1
GPGPU computing performance is a big part of modern GPU design, especially in workstation environments. LuxMark is a long-standing OpenCL standard, based on the LuxRender tool and provides a good view of how different GPU architectures compare to typical OpenCL workloads. Today we are testing our field of graphics cards in the most intensive computing scenes, hotels.
Looking at the LuxMark Hotel scene, we will have more competitive results than expected. While AMD’s GPUs traditionally have an advantage over NVIDIA’s options in OpenCL, based on highly rated GPU computing performance. About 8.5 TFLOPS of an AMD Fiji XT GPU is on par with the P5000 having 8.5 TLOPS.
OpenCL has the ability to balance with many computing devices, which means a very impressive score for Radeon Pro Duo when both GPUs are being used.
where can you get a NVIDIA Quadro P5000 online
PNY Quadro P5000 VCQP5000-PB 16GB 256-bit GDDR5x PCI Express 3.0 X16 Full Height Video Card – Workstation: Buy it now
PNY NVIDIA Quadro P5000 Graphics Card: Buy it now
HP Quadro P5000 Graphic Card – 16 GB GDDR5X – Dual Slot Space Required: Buy it now
Blender 2.78b
Blender is a popular open source project for 3D modeling and animation. Blender supports both OpenCL and CUDA pipelines, allowing interesting comparison between GPU vendors.
It is important to note that we tested Blender with the default settings. This means that tile sizes do not necessarily have to be optimized for the specific GPU that we are testing. Additional performance should be possible by experimenting with tile sizes for each specific GPU.
Looking at the results, it seems that the CUDA kernel line has the advantage of the OpenCL code path. Using OpenCL, Radeon Pro Duo only offers a modest improvement with 16.38 TFLOPS computing performance in dual GPUs compared to a Quadro P5000.
3DS Max (NVIDIA Iray +)
3DS Max is a professional 3D model, animation and visual display. Due to its popularity, there are many display tools available for use in 3ds Max. In this test, we are considering the Iray + tool developed by NVIDIA. This rendering application uses CUDA, and cannot run on AMD Radeon Pro Duo (it can run exclusively on CPUs of AMD GPUs).
We want to test Iray to get an idea of the relative performance between Quadro options looking for it in a real world application. As you can see, even the $ 400 Quadro P2000 provides a greater performance improvement than X-ray detection on the 8-core Intel Core i7-5960X CPU.
While the Quadro P4000 makes the scene 59% faster than the P2000, the performance up to P5000 is only 15%. Although these results will vary depending on the workload, it shows that the diminishing returns are possible when you increase GPU capacity.
Conclude
For readers who use the consumer graphics market, these prices seem very high. One important thing to remember is that, while they can share the same physical hardware, the software side for professional graphics cards is different.
For professional applications, stability and compatibility are key. Additional work is included in drivers for professional graphics cards to ensure that they will perform every day for users who are making a living from the applications they are helping to accelerate. And for these consumers, the price increase is worthwhile.
As you can see from our benchmark results, in the professional graphics market, the choice of graphics card reduces the applications you are using.
The important things to consider are what API implements the applications you use most often (OpenCL, CUDA, OpenGL, …), the ability of these applications to take advantage of many computing devices, as well as targeted system requirements (such as energy consumption and physical shape).
Obviously, if you are using applications that focus exclusively on OpenCL, AMD offers attractive options in the workstation graphics card market. Although the starting price of $ 1500 for Radeon Pro Duo is difficult to swallow about the performance level it offers, the current $ 800 price is attractive to users looking for very high OpenCL visibility, as long as Your application will support Multiple target OpenCL devices.
For external CUDA applications or instead of OpenCL, NVIDIA’s Quadro GPU doesn’t disappoint you. In Blender, where we can test OpenCL and CUDA, it’s clear that CUDA provides a performance advantage when combined with one of the Quadro Pascal we tested. While the Quadro P5000 may not be the absolute winner, it is very close to Radeon Pro Duo while running only on a physical GPU.
We are delighted to see AMD’s last FirePro products featuring high-end Vega GPU features, but now, NVIDIA offers the highest-performance workstation graphics cards with their latest Quadro products. Pascal architecture provides a platform that NVIDIA can build more efficient graphics cards for professional users who are more aware of their size, power and temperature at the desk. Users are pushing the limits of rendering and calculations will do well with one of these graphics cards.