Cisco Blogs


Cisco Blog > Data Center

From desktop to Data Center…changing the game with graphics

What if you were able to give everyone in your organization the flexibility and freedom to securely work anywhere in the world and on any device? What types of productivity gains would your company see as a result? What efficiencies or cost savings might your IT department receive from moving desktops to your datacenter and managing these virtual workstations through one single pane of glass?

social-03 

Our Cisco UCS team is excited to present the new Maxwell generation NVIDIA Tesla M6 GPU for the Cisco UCS B200M4 Blade and the NVIDIA Tesla M60 GPU for Cisco 2U Rack Rack Servers. Cisco and NVIDIA have joined forces to deliver this new graphics solution.  Combining security, reliability and manageability from Cisco UCS and adding NVIDIA’s GRID technology, we’re able to deliver performance and speed needed to run high-end applications on virtual desktops. What’s better is that you have two form factor options to fit your organization’s’ data center footprint. 

The M60 Rack GPU is supported with UCS Manager 3.1(1) and later & Cisco Integrated Management Controller (CIMC) 2.0(9) and later; the M6 Blade GPU is supported with both UCS Manager 3.1(1) and 2.2(7) and later.

 

With Cisco UCS & NVIDIA GRID, you can now expand your virtualization footprint without compromising performance or user experience while also increasing security. This means, you can empower your workforce to create anything around the world, from any location with the ease and flexibility.

 

The new Cisco & NVIDIA M6 Blade GPU solution is fully integrated with the flagship B200M4 Server, supporting all CPU configurations and performance on par with NVIDIA K2 GPU, at less than half the power profile!  With two of the M60 GPU’s on Cisco UCS C240M4, you can now enable high density NVIDIA Tesla Compute and GRID 2.0 VDI user consolidation, which is over 8,000 CUDA cores & 32GB of GPU memory for up to 64 GPU-accelerated virtual desktop users.
nvidia-grid-2-0-1

This is especially exciting for organizations in the Oil and Gas, Manufacturing and Design industries since historically, this type of work demanded high-end applications like ESRI, AutoCAD, Petrel and Siemens to be used in on-site with workstations. But now, these types of applications can be powered virtually through your data center to any device.

Take an airplane manufacturer for example. With a follow-the-sun working model, this organization can now empower its employees to design from anywhere in the world with the device they prefer, while all working on the same application in real-time. This helps the company save time, money and fuels rapid innovation. Last but not least, this is all done while maintaining security for the organization.

Empower your employees to create flexibly, securely and from any device and any location in the world. Now possible with Cisco UCS & NVIDIA GRID.   

social-04

Productivity, mobility, security, and flexibility for all. This changes everything.

Learn how others are transforming their data centers with Cisco UCS.

Tags: , , , , , , , , , , , , , , ,

Agile Application Delivery with UCS-Powered VDI

vmworld-hero-960x800

We’re off to another edition of VMworld with some great technology to showcase. Our theme this year is your data center is everywhere, the center of innovation. And talking about innovation we will be demoing some recent work around our desktop virtualization solution with VMware. We will have demo pods showcasing our UCS solutions for Virtual SAN and VDI as well as a demo showing graphic acceleration for VDI with Cisco UCS and NVIDIA GRID on Horizon. By the way, did you catch NVIDIA’s announcement today about GRID 2.0. This is pretty exciting news as graphic acceleration is key to a seamless user experience when you virtualize desktops or applications. With Cisco UCS a market leader in blade servers we’re looking forward to bringing virtualized GPUs to our B-series servers. As we see more and more graphic-intensive applications being virtualized to protect intellectual property by keeping data safely in the data center Cisco is working with NVIDIA and ISVs to qualify hardware configuration for specific applications. Read about Esri ArcGISPro recent testing on Cisco UCS C240-M4 and NVIDIA K2 GPU. Read More »

Tags: , , , , , ,

Modern GPU Integration in MPI

Today’s guest post is from Rolf vandeVaart, a Senior CUDA Engineer with NVIDIA.

GPUs are becoming quite popular as accelerators in High Performance Computing clusters. For example, check out Titan; a recent entry into the Top 500 list from Oak Ridge Laboratories. Titan has 18,688 nodes (299,008 CPU cores) coupled with 18,688 NVIDIA Tesla K20 GPUs.

To help ease the programming burden working with GPU memory in MPI applications, support has been added to several MPI libraries such that the MPI library can directly send and receive the GPU buffers without the user having to stage them in host memory first. This has sometimes been referred to as “CUDA-aware MPI.”

Read More »

Tags: , , ,