You are Home   »   News   »   View Article

Working on subsurface data remotely - NetApp

Friday, October 16, 2015

Data management company NetApp has joined together with networking company Cisco, microchip manufacturer NVIDIA, and remote desktop company Citrix, to build a system enabling you to work on remotely stored subsurface data, using just a tablet computer.

Based on a paper written by Osama Qazi, global energy architect with NetApp, and Doug Wycoff, systems engineer, Cisco Systems

Allowing a mobile workforce to visualize the latest subsurface data, wherever they happen to be, is becoming an essential part of upstream oil and gas processes.

Such visualization requires either local dataset copies, which can be difficult and potentially risky to provide, or remote visualization.

Up to now, remote visualization solutions have been complicated to deploy and manage and haven't provided the quality or responsiveness that users expect.

NetApp and Cisco partnered with Citrix and NVIDIA to address the demanding visualization needs of upstream oil and gas.

This new solution, FlexPod Datacenter with Citrix XenDesktop and NVIDIA GRID, was architected to address the challenges of 21st-century exploration and production.

It results in a solution that makes visualization accessible to everyone who needs it-including users with thin devices such as tablets.

A single, reliable infrastructure is able to address a full spectrum of visualization needs - those of local users, visualization rooms, remote office users, and workers in the field.

The upstream environment

The upstream environment has needs that are significantly different from those of typical IT infrastructures.

It's not uncommon to run hundreds of different applications with a broad mix of I/O needs that can include both sequential and random data access, extremely large file sizes, and huge numbers of smaller files.

Workflows continue to evolve because of the rapid growth in dataset size and the need to integrate diverse, multidisciplinary data types to more accurately understand the earth's subsurface.

Feeding all this data to visualization workstations outside the data centre has become increasingly difficult.

Network connections lack the necessary bandwidth for real-time operations and workflows increasingly span organizational boundaries.

Continued reliance on visualization workstations is creating significant challenges.

Copying data to a workstation to facilitate visualization, which is sometimes the only option, wastes valuable time while transfers take place.

Data on local workstations may not be protected by backup or disaster recovery, risking data loss.

Storing data on workstations outside the data centre increases security risks.

Moving data outside the data centre or across international boundaries may violate regulations and corporate governance requirements.

Workstation failures can have a big impact on the productivity of geologists, geophysicists, engineers, and other visualization users.

Software licensing costs per seat can be extremely high, while productivity per license may be low.

Collaboration, both local and remote, often requires additional data movement, creating delays of hours or even days before collaboration can begin.

Bringing a new workstation online to support a new user can take weeks.

Setting up workstations for workers in remote locations and supplying them with data creates additional challenges and additional expense.

The upshot is that significant opportunities exist to increase productivity, enhance collaboration, decrease risks, simplify the overall IT environment, and decrease costs.

The NetApp solution

The FlexPod Datacenter with Citrix and NVIDIA solution brings together best-in-class components to overcome the challenges just discussed.

Data is centralized in the data centre, eliminating the need for data copies, reducing risk and decreasing management overhead, and increasing productivity.

Anyone who needs visualization can request a session and begin working in a matter of moments.

FlexPod Datacenter with Citrix and NVIDIA extends the familiar concept of virtual desktop infrastructure (VDI) to encompass real-time 3D graphics capabilities.

An authorized user simply requests a desktop or application session. Data is processed and rendered within the data centre, and the results of visualization are transmitted over the network to the user.


FlexPod is a proven data centre solution from NetApp and Cisco, with a flexible, shared infrastructure that easily scales to support growing workload demands without impacting performance.

FlexPod has been widely used in virtual desktop infrastructure (VDI) deployments for the last four years.

It offers faster deployment with less risk. The pre-validated design means that you can get FlexPod up and running quickly and deploy applications faster with fewer problems.

It offers the ability to scale up and out. FlexPod grows and adapts to meet your changing needs.

It offers investment protection. The scalability and flexibility of FlexPod give the solution a longer life. You can easily repurpose components to address changing needs.

Also NetApp, Cisco, and Citrix partnered to provide cooperative support. This approach resolves 98 per cent of support issues on first contact.


The NetApp FAS8000 storage hardware reduces overall storage costs while delivering the necessary I/O performance for virtual desktop infrastructure (VDI) and demanding geology and geophysics applications.

It supports both all-flash and hybrid storage configurations.

To deliver the necessary throughput and low latency required for VDI, all-flash configurations are typically used.

All-flash FAS nodes can be combined in the same scale-out cluster with hybrid (HDD plus flash) storage nodes suitable for storing large datasets with high throughput.

You can also replicate or back up your VDI environment (from expensive SSD storage) to more economical HDD storage on other nodes.


Cisco UCS C240 M4 rack servers are typically chosen for the FlexPod Datacenter with Citrix and NVIDIA solution.

These servers feature extended memory for faster rendering, support of bigger datasets, more desktops per server, and low latency.

This server supports up to two double-wide GPUs in a slim 2U form factor. GPUs can be used to perform parallel tasks (GPU computing) as well as visualization.

Graphics processing

NVIDIA GRID technology offloads graphics processing from the CPU to the GPU in virtualized environments.

GRID GPUs are specifically designed for use in virtual environments. The cards lack external video connections.

NVIDIA GRID supports two modes of operation, 'GPU pass-through' and 'Virtual GPU' (vGPU).

With GPU pass-through, each user is assigned a physical GPU.

With Virtual GPU (vGPU), a single physical GPU is shared by up to eight users.

GPU pass-through is well suited to power users such as geologists and geophysicists, while vGPU is optimal for less intensive graphics users.

The FlexPod Datacenter solution with Citrix and NVIDIA uses the NVIDIA GRID K2, featuring two high-end Kepler GPUs with a total of 3,072 CUDA cores and 8GB of GDDR5 memory.

Because each Cisco UCS C240 M4 server supports two GRID K2 devices (4 GPUs), a single 2U server is capable of supporting up to four users with GPU pass-through or up to 32 users with vGPU.


Citrix XenDesktop delivers applications and desktops as secure mobile services.

XenDesktop with HDX 3D Pro delivers a native, touch-enabled experience that is optimized for the type of device, as well as the network.

XenDesktop was among the first to support the use of GPUs for 3D graphics in virtualized environments and the first hypervisor to support NVIDIA GRID in production.

Overall oil and gas benefits

Here are some benefits specific to the upstream oil and gas industry.

Geoscientists, engineers, and business decision makers can see important results in near real time, without the bottlenecks that result from transferring huge datasets over network connections or by mail.

Collaborators view and manipulate the same images, eliminating potential points of confusion and miscommunication and saving valuable time.

You can eliminate the expense and complexity of dedicated workstations by replacing them with a much more efficient and scalable shared resource capable of supporting workers wherever they happen to be.

You can have new infrastructure up and running in less time with less effort.

Because FlexPod Datacenter is well established, many resources exist to facilitate and streamline deployment. The combination of FlexPod and XenDesktop is well understood.

You can size your UCS servers to address the needs of power users using GPU pass-through or shared GPU seats using vGPU.

Associated Companies
» NetApp
comments powered by Disqus


To attend our free events, receive our newsletter, and receive the free colour Digital Energy Journal.


Clustering Considerations in the Machine Learning Workflow – Examples with Exploration Data
Philip Lesslar
from Precision DM


Latest Edition Apr-May 2020
Apr 2020

Download latest and back issues


Learn more about supporting Digital Energy Journal