Technical Support

User Rating:  / 0
AddThis Social Bookmark Button
Change letter size:

If you need technical support for installing or using the rCUDA remote GPU virtualization middleware, please do not hesitate to contact us. You can find information about how to contact us in the "Contact us" tab of this web page. On the other hand, you can also access our FAQ list as well as the many documents available about rCUDA. Any other inquires about rCUDA are also welcome.

- Frequently asked questions (FAQs), where we will be adding more issues consulted by our users.

- A list of papers and presentations with technical information about rCUDA.

Documentation

User Rating:  / 11
AddThis Social Bookmark Button
Change letter size:

rCUDA User's Guide and Quick Start Guide

- The rCUDA User's Guide for current rCUDA release can be found here.

- You can find the Quick Start Guide here.

White Papers

- The rCUDA middleware and applications. Publication date December 2016
- Deploying rCUDA in cloud computing environments. Publication date December 2016
- rCUDA Frequently Asked Questions. Publication date December 2016

Papers

- C. Reaño, Matthew J. Leslie, and F. Silla, schedGPU: Fine-Grain Dynamic and Adaptative Scheduling for GPUs. In the 2016 International Conference on High Performance Computing & Simulation (HPCS 2016), Innsbruck, Austria, July, 2016

- C. Reaño and F. Silla. Reducing the Performance Gap of Remote GPU Virtualization with InfiniBand Connect-IB. In the 21st IEEE Symposium on Computers and Communications (ISCC 2016), Messina, Italy, June, 2016

- F. Pérez, C. Reaño, and F. Silla. Providing CUDA Acceleration to KVM Virtual Machines in InfiniBand Clusters with rCUDA. In the 16th IFIP International Conference on Distributed Applications and Interoperable Systems (DAIS 2016), Heraklion, Crete, Greece, June 2016

- S. Iserte, J. Prades, C. Reaño, and F. Silla. Increasing the Performance of Data Centers by Combining Remote GPU Virtualization with Slurm. In the 16th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGRID 2016), Cartagena, Colombia, May 2016

- F. Silla, C. Reaño, J. Prades, S. Iserte. Benefits of remote GPU virtualization: the rCUDA perspective. In GPU Technology Conference (GTC), San Jose, CA, USA, April 2016

- J. Prades, C. Reaño, and F. Silla. CUDA Acceleration for Xen Virtual Machines in InfiniBand Clusters with rCUDA. In 21st ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, Barcelona, Spain, March 2016

- F. Silla, J. Prades, S. Iserte, and C. Reaño. Remote GPU Virtualization: Is It Useful?.  In Proceedings of the 2nd IEEE International Workshop on High-Performance Interconnection Networks in the Exascale and Big-Data Era (HiPINEB 2016), held in conjunction with IEEE HPCA 2016. Barcelona, Spain. March 2016

- C. Reaño, F. Silla, G. Shainer, and S. Schultz. Local and Remote GPUs Perform Similar with EDR 100G InfiniBand.  In Proceedings of the ACM Middleware Conference 2015. Vancouver, BC, Canada. December 2015

- Carlos Reaño and Federico Silla. A Performance Comparison of CUDA Remote GPU Virtualization Frameworks. In Proceedings of the IEEE Cluster 2015 Conference. Chicago, IL, USA, September 2015

- Carlos Reaño and Federico Silla. InfiniBand Verbs Optimizations for Remote GPU Virtualization. In Proceedings of the 1st IEEE International Workshop on High-Performance Interconnection Networks Towards the Exascale and Big-Data Era (HiPINEB), held in conjunction with IEEE Cluster 2015. Chicago, IL, USA. September, 2015

- Carlos Reaño, Federico Silla, Adrián Castelló, Antonio J. Peña, Rafael Mayo, Enrique S. Quintana-Ortí, and José Duato. Improving the user experience of the rCUDA remote GPU virtualization framework. In Concurrency and Computation: Practice and Experience, Volume 27, Issue 14, pages 3746-3770,  September 2015

- Blesson Varghese, Javier Prades, Carlos Reaño, and Federico Silla. Acceleration-as-a-Service: Exploiting Virtualised GPUs in HPC Clusters for an Actuarial Science Application. In Proceedings of the 11th IEEE International Conference on eScience. Munich, Germany, August 2015

- Carlos Reaño, Ferran Pérez, and Federico Silla. On the design of a demo for exhibiting rCUDA. In Proceedings of the 15th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGRID). Shenzhen, Guangdong, China. May 2015.

- Antonio J. Peña, Carlos Reaño, Rafael Mayo, Enrique S. Quintana-Ortí, and José Duato. A complete and efficient CUDA-sharing solution for HPC clusters. In Parallel Computing Journal, Volume 40, Issue 10, pages 574–588. December 2014.

- Sergio Iserte, Adrián Castelló, Rafael Mayo, Enrique S. Quintana-Ortí, Federico Silla, José Duato, Carlos Reaño, and Javier Prades. SLURM Support for Remote GPU Virtualization: Implementation and Performance Study. In Proceedings of the 26th International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD 2014). Paris, France, October 2014.

- Carlos Reaño, Federico Silla, Antonio J. Peña, Gilad Shainer, Scot Schultz, Adrian Castelló, Enrique S. Quintana-Ortí, and José Duato. POSTER: Boosting the performance of remote GPU virtualization using InfiniBand Connect-IB and PCIe 3.0. In Proceedings of the IEEE Cluster 2014 Conference, Madrid, Spain, September 2014.

-Carlos Reaño, Rafael Mayo, Enrique S. Quintana-Ortí, Federico Silla, José Duato and Antonio J. Peña. Influence of InfiniBand FDR on the Performance of
Remote GPU Virtualization
. In Proceedings of the IEEE Cluster 2013 Conference, Indianapolis, IN (USA), September 2013.

- Carlos Reaño, Antonio J. Peña, Federico Silla, José Duato, Rafael Mayo, and Enrique S. Quintana-Ortí. CU2rCU: Towards the complete rCUDA remote GPU virtualization and sharing solution. In Proceedings of the 2012 International Conference on High Performance Computing (HiPC 2012), Pune, India, December 2012.

- José Duato, Antonio J. Peña, Federico Silla, Juan C. Fernández, Rafael Mayo, and Enrique S. Quintana-Ortí. Enabling CUDA acceleration within virtual machines using rCUDA. In Proceedings of the 2011 International Conference on High Performance Computing (HiPC 2011), Bangalore, India, December 2011.

- José Duato, Antonio J. Peña, Federico Silla, Rafael Mayo, and Enrique S. Quintana-Ortí. Performance of CUDA virtualized remote GPUs in high performance clusters. In Proceedings of the 2011 International Conference on Parallel Processing (ICPP 2011), Taipei, Taiwan, September 2011.

- José Duato, Antonio J. Peña, Federico Silla, Rafael Mayo, and Enrique S. Quintana-Ortí. rCUDA: reducing the number of GPU-based accelerators in high performance clusters. In Proceedings of the 2010 International Conference on High Performance Computing and Simulation (HPCS 2010), pages 224-231, Caen, France, June 2010.

- José Duato, Francisco D. Igual, Rafael Mayo, Antonio J. Peña, Enrique S. Quintana-Ortí, and Federico Silla. An efficient implementation of GPU virtualization in high performance clusters. In Euro-Par 2009, Parallel Processing - Workshops, volume 6043 of Lecture Notes in Computer Science, pages 385-394. Springer-Verlag, 2010.

Presentations

The rCUDA technology has been presented in several conferences and exhibitions. You can find below the current list of presentations made during the last years. The PDF file of each presentation is also available. Please, notice that rCUDA is a live project and, therefore, it evolves over time. Thus, some of the information contained in the presentations below may be obsolete now. The list is just maintained as a historical record.

- Federico Silla. The remote GPU virtualization from the rCUDA point of view. In HPC Knowledge Meeting'16, Barcelona, Spain, April 21, 2016. You can also access a video with the presentation in this link.

- Federico Silla. Benefits of remote GPU virtualization: the rCUDA perspective. In GPU Technology Conference 2016 (GTC'16), San Jose (CA), USA, April 2016.

- Federico Silla. Increasing cluster performance by combining rCUDA with Slurm. In HPC Advisory Council Switzerland Conference 2016, Lugano, Switzerland, March 2016. You can also access a video with the presentation in this link.

- Federico Silla. Benefits of remote GPU virtualization. In The International Conference for High Performance Computing, Networking, Storage and Analysis (SC15), Mellanox Booth, Austin, TX, U.S.A., November 2015.

- Federico Silla. Remote GPU virtualization. Opening talk in Jornadas Sarteco 2015, Cordoba, Spain, September 2015.

- Federico Silla. Is remote GPU virtualization useful?. In HPC Advisory Council Spain Conference 2015, Barcelona, Spain, September 2015. You can also access a video with the presentation in this link.

- Federico Silla. Increasing cluster throughput with Slurm and rCUDA. In Slurm User Group Meeting 2015, Washington DC, USA, September 2015.

- Federico Silla. Remote GPU virtualization: pros and cons of a recent technology. In HPC Advisory Council Brazil Conference 2015, Petropolis, Brazil, August 2015.

- Federico Silla.  The rCUDA technology: an inexpensive way to improve the performance of GPU-based clusters. Talk at Computer Engineering Laboratory, Delft University of Technology. Delft, Netherlands. April 2015 

- Federico Silla. Increasing Cluster Throughput while Reducing Energy Consumption for GPU Workloads. In HPC Advisory Council Switzerland Conference 2015, Lugano, Switzerland, March 2015. You can also access a video with the presentation in this link.

- Federico Silla. Reducing Costs with GPU Virtualization. In The International Conference for High Performance Computing, Networking, Storage and Analysis (SC14), Mellanox Booth, New Orleans, LA, U.S.A., November 2014.

- Federico Silla.  rCUDA: Virtualizing GPUs to reduce cost and improve performance. In STAC Summit London 2014, London, UK, October 2014.

- Federico Silla. Increasing the throughput of your GPU-enabled cluster with rCUDA. In HPC Advisory Council Spain Conference 2014, Santander, Spain, September 2014. You can also access a video with the presentation in this link.

- Sergio Iserte.  Extending Slurm with Support for Remote GPU Virtualization. In Slurm User Group Meeting 2014, Lugano, Switzerland, September 2014.

- Federico Silla. rCUDA: share and aggregate GPUs in your cluster. In The International Conference for High Performance Computing, Networking, Storage and Analysis (SC13), Mellanox Booth, Denver, CO, U.S.A., November 2013.

- Federico Silla. rCUDA: towards energy-efficiency in GPU computing by leveraging low-power processors and InfiniBand interconnects. In HPC Advisory Council Spain Conference 2013, Barcelona, Spain, September 2013. You can also access a video with the presentation in this link.

- Rafael Mayo. rCUDA: a ready-to-use remote GPU virtualizacion framework. In HPC Advisory Council European Conference 2013, Leipzig, Germany, June 2013.

- Federico Silla. Improving the Efficiency of your GPU-accelerated Cluster with rCUDA. In The International Conference for High Performance Computing, Networking, Storage and Analysis (SC12), Mellanox Booth, Salt Lake City, U.S.A., November 2012.

- Todd Wilde and Rafael Mayo. Maximize the GPU Performance in Your Compute Cluster Using rCUDA Virtual GPU Technology. Webminar: http://www.mellanox.com/webinars/2012/Using-rCUDA-Virtual-GPU-Technology, 2012.

- Antonio J. Peña and Rafael Mayo. rCUDA 4: GPGPU as a service in HPC clusters. In HPC Advisory Council Spain Conference 2012, Málaga, Spain, September 2012.

- Federico Silla and Antonio J. Peña. rCUDA, an approach to provide remote access to GPU computational power. In HPC Advisory Council Switzerland Conference 2012, Lugano, Switzerland, March 2012.

- Federico Silla. Improving the Efficiency of your GPU-accelerated Cluster with rCUDA. In The International Conference for High Performance Computing, Networking, Storage and Analysis (SC11), Mellanox Booth, Seattle, U.S.A., November 2012.

- Rafael Mayo. rCUDA, an approach to provide remote access to GPU computational power. In HPC Advisory Council China Workshop 2011, Jinan, China, October 2011.

- Rafael Mayo. rCUDA, an approach to provide remote access to GPU computational power. In HPC Advisory Council European Workshop 2011, Hamburg, Germany, June 2011. 

Tutorials

- Federico Silla and Carlos Reaño. Reducing Power Consumption of Data Centers with rCUDA. In HiPEAC 2016 Conference. Prague, Czech Republic, January 2016.

- Federico Silla and Carlos Reaño. Improving overall performance and energy consumption of your cluster with remote GPU virtualization. In Middleware 2015 Conference, Vancouver, Canada, December 2015.

- Federico Silla and Carlos Reaño. Make your heterogeneous cluster more flexible with remote GPU virtualization. In The 24th International Conference on Parallel Architectures and Compilation Techniques (PACT'15), San Francisco, CA, U.S.A., October 2015.

- Federico Silla and Carlos Reaño. On the use of remote GPU virtualization for managing the GPUs of your cluster in a flexible way. In IEEE Cluster 2014 Conference, Madrid, Spain, September 2014.

FAQs

User Rating:  / 7
AddThis Social Bookmark Button
Change letter size:

Q.- How to install rCUDA?

A.- Steps to install rCUDA binaries

1.                  Uncompress the rCUDA package

2.                  Copy the rCUDAd folder to the server node (with GPU)

3.                  Copy the rCUDAl folder to the client(s) node(s) (without GPU)

Steps to use the rCUDA samples (Server side):

 1.                  Run the rCUDAd executable at rCUDAd folder (See the options at rCUDA_guide in doc folder)

 

Steps to use the CUDA samples (in the client side):

 

1.                  Install the NVIDIA sdk

 

2.                  Go to NVIDIA_CUDA_x.x_Samples folder and compile it using EXTRA_NVCCFLAGS=-cudart=shared

 

3.                  Export the rCUDA environment variable RCUDA_DEVICE_COUNT set to the amount of GPUs available to the application

 

4.                  Export as many rCUDA environment variables RCUDA_DEVICE_X as GPUs are provided to the application in previous step (notice that the "X" in the variable is the GPU number). Each RCUDA_DEVICE_X variable must follow the syntax <server[@<port>]>[:GPUnumber] (for example RCUDA_DEVICE_0=192.168.0.1, or RCUDA_DEVICE_1=192.168.0.1:1)

 

5.                  Update the LD_LIBRARY_PATH  environment variable with the path where the rCUDAl folder is (typically /$(HOME)/rCUDAl )

 

6.                  Make sure to have the rCUDA server daemon running

 

7.                  Finally you can execute the CUDA samples


Q.- While trying to compile a source code, I'm getting the error “undefined reference to cuXxxxxxXxxxxx”.

A.- “cuXxxxxxXxxxxx” are CUDA Driver API calls. Currently rCUDA only supports Runtime API calls.

 

Q.- I'm not able to run applications in rCUDA. I'm getting initialization errors.

A.- Check:

  1.- TCP/IP connection between host and guest, including firewall settings

  2.- On the CLIENT, export RCUDA=x.x.x.x, where x.x.x.x is the IP of the rCUDA server. Server names will also work.

  3.- On the CLIENT, LD_LIBRARY_PATH is pointing to the path of the rCUDA library, (e.g., /home/user/rCUDA/framework/rCUDAl or /usr/local/cuda/lib64 if you replaced the original one)

4.- Status of the rCUDA daemon on the host. You can see logs usually on /var/log/messages. In addition, for testing purposes it is usefull to run the rCUDA server in interactive and/or verbose mode: ./rCUDAd -iv

 

Q.- I'm getting the error “Error while loading shared libraries: libcudart.so.4.1: cannot open shared object file: No such file or directory.”

 A.- This is a run-time error, stating that your LD_LIBRARY_PATH is not pointing where the rCUDA library is.

 

Q.- I'm not able to include rCUDA_util.h.

A.- That header file is not striclty needed to compile rCUDA applications, although offers a convenient function to setup kernel launch arguments in fewer lines of code. This is a compilation problem, solved by just adding the -I flag to the compiler: -I path/to/rCUDA_util.h (e.g., /home/user/rCUDA/util).

 

Change letter size:

Gold Sponsors

Silver Sponsors

Logo gva

 

logo bright

logo nvidia