The world has embraced hybrid work but not all workforces have managed to adjust to this new model of working with ease. For data scientists, setting up an environment or software stack can be cumbersome and disruptive, taking time and effort away from doing their jobs.
At the GTC webinar entitled Accelerate Your Data Science Workflows From Anywhere, experts from HP, Alex Duncan, Product Manager, Data Science Hardware & OS, and Mirela Cunjalo, HP Anyware Lead Product Manager, shared insights into how data scientists can use the Z by HP Data Science Stack Manager and HP Anyware to boost their workflows and spend more time doing their best work from wherever they are.
You can read highlights from the session below.
The digital world generates tons of data. Whether you’re working in education, finance, the government sector, M&E, or another industry, data is the key to business growth.
Understanding the meaning of that data in context to one’s company and applying that knowledge to develop one’s business is crucial to long-term growth.
Data scientists examine and act on these data, statistics, and graphical models. They are an essential cog in the corporate wheel and can include analysts, researchers, developers and business managers.
During the GTC webinar on data science workflows, Alex Duncan prefaced the unique challenges data scientists face in a hybrid environment by explaining the demands they encounter. “Data science environments, colloquially known as a data science software stack, contain many different software tools that are integrated together for both specific and general workflows.”
In the hybrid world we live in now, IT teams have to manage employees who are either 100% remote, 100% in the office, a combination of these, or working from somewhere else entirely. According to an HP study, 42% of data scientists say that configuring their environment in such instances takes too much time.
As such, IT and end users encounter challenges in the following areas:
Select Z by HP devices come equipped with pre-loaded data science tools that have already been tested and set to update automatically. With the Data Science Stack Manager, it’s easier for users to customize their environments for projects and workflows.
As Duncan elucidated, the stack manager is a workstation OEM-developed graphical user interface for customizing data science software and allows users to spend less time on tasks such as installing and managing the environment and workflows, and more time on completing data science projects. The stack manager is available on Ubuntu and Microsoft WSL 2.
Another benefit of the Data Science Stack Manager is the graphical interface which makes the system simple for users to understand and start working on. For hybrid workers, the Z by HP data solutions can interface with numerous cloud providers, such as AWS, Google Cloud Provider and Microsoft Azure. For added security and mobility, the HP Anyware remote access software can be installed in a hybrid environment so data scientists can work wherever they are.
Mirela Cunjalo spoke to the power of HP Anyware for digital workspaces as a software that helps data scientists work efficiently. Based on PC-over-IP (PCoIP®) technology, Anyware combines the best features of ZCentral Remote Boost and Teradici CAS. Both software have the distinction of winning Emmy® Engineering Awards.
HP Anyware includes three components:
“HP Anyware provides IT a lot of flexibility around infrastructure and how they support end-users,” Cunjalo explained. “A big part of improving flexibility is to enable the end-user to efficiently work even when they’re not in the office.”
Data science workflows can be very compute-intensive as large amounts of variable data need to be sorted and analyzed. When a data scientist is working in the office or at home on a big workstation, it’s easier to compute large numbers.
But a user can’t carry a heavy workstation with them. Most likely, they’ll take a lighter device, such as a lightweight laptop if they’re on the go. In these scenarios, a data scientist can log into HP Anyware on their client device to remotely connect to the data center from wherever they have network access.
Even in an office environment, HP Anyware allows the user and IT teams flexibility. Since Anyware can be installed on almost any device to access information from the host desktop, if an endpoint fails, the work is still safe and accessible.
HP Anyware software can be deployed in a variety of environments, public or private cloud, such as AWS, Microsoft Azure, or Google Cloud Provider. It can also be deployed on on-premises data centers, edge, virtualized environments, physical workstations that run Windows®, or MacOS®, Linux®, RHEL, Ubuntu, Rocky, CentOS, or any combination of these.
The software provides flexibility in terms of operating systems, as well. Hosts can be installed on Windows®, Linux®, or MacOS® for client devices running Windows, Linux, MacOS, Chrome OS®, iOS®, or Android®. Anyware supports numerous devices, from a PC, Mac®, laptop, Chromebook®, tablet, to integrated monitors, Zero Clients, and Thin Clients.
“When an end-user is remote, not physically in the office, security is the key factor that IT has to consider,” said Cunjalo during the webinar. HP Anyware is powered by the PCoIP protocol, as such data never leaves the data center or the office because Anyware only sends AES 256 encrypted image pixels to the user.
Instead of downloading heavy files or sensitive information onto a local system that decreases productivity and increases risks of cyberattacks, data scientists can use HP Anyware to connect to their remote desktop wherever it is located. The information they access will be transferred as image pixels, not raw data, to the end-user's device, and will not need to be saved on the endpoint, thus preventing the loss or theft of information.
Additionally, while many businesses that work in remote or hybrid settings use VPNs, with HP Anyware, a VPN is not required as the software is already in-built with multi-factor authentication protocols.
Wherever an end-user is, remote or in the office, they need high-performance to be productive. “A big part of productivity is being able to interact and work with applications on a daily basis, and work with them well,” emphasized Cunjalo. If a remote desktop doesn’t offer similar levels of responsiveness as a local machine, a user will struggle to be productive.
Since HP Anyware provides high color-accuracy and text clarity with peripherals, data scientists won’t encounter distortions on the remote desktop.
HP Anyware’s Auto-Offload feature automatically switches between CPU and GPU resources depending on changing network conditions and available bandwidth, so the user experience isn’t impacted. This also means that IT doesn’t have to spend time managing the end user’s environment which varies according to their network and workloads.
One of the biggest strains on hybrid businesses has been the loss of over-the-shoulder collaboration. Users are dispersed in various locations, not just in the same city but sometimes across the globe. Being so far away from colleagues makes it difficult for data scientists to collaborate and complete projects. Additionally, sharing information with colleagues can mean uploading and downloading large files, that takes time and uses up bandwidth.
HP Anyware’s collaboration feature allows users to invite one another to a remote desktop session to talk through data points, designs, and applications, as if they were physically in the same location. They will both be able to see and collaborate on whatever application or content is on that desktop, with almost the same type of performance that one would get if one was sharing a room.
Both users will have the option to take control of the keyboard or mouse so that collaboration feels all the more seamless. Since the color and text accuracy and clarity are the same for both users in a collaborative environment, data scientists won’t feel like they’re in two different environments.
Additionally, since all the data is in the data center that both data scientists are remotely accessing, there is no need to download files to share between each other.
The Z by HP Data Science Stack Manager and HP Anyware combination provides data scientists the flexibility to work anywhere with a network connection, using almost any type of device, with the knowledge that their work and the business information they access remains safely in the data center.
Z by HP and Anyware provide the ability to collaborate while in different locations and the use of any peripherals to be productive so data scientists can work proficiently in any type of hybrid work environment.
Learn more about Z by HP Data Science Stack Manager, along with a demonstration of how to download and install the software, and see how HP Anyware can improve data science workflows where users are. Watch the complete session here.