Using Workstations To Reshape Your Artificial Intelligence Infrastructure

How Workstations Can Play A Pivotal Role in Supporting Your AI Strategy

Today, artificial intelligence (AI) applications are reshaping large and small companies’ products, services, and business models. Decision makers need to make critical hardware and software choices to achieve AI strategy success. As smaller firms to global enterprises assess their business cases and work flows the need for highly-scaled servers and cloud platforms are no longer the only option to build a successful AI infrastructure.

The study shows that firms are already using workstations to lower the cost, increase the security, and speed up their AI infrastructure. The addition of workstations into a firms AI workflow allows servers and cloud platforms to be tasked with business cases that require more robust computing while workstations take on tasks with longer timeframes and smaller budgets.

  • 6% are using workstations as to run core AI applications today
  • 32% are using workstations to train or process algorithms
  • 26% are using workstations as to run core AI applications today

AI is a focus for the future

Firms need to invest in the right platforms to make their artificial intelligence work best for their business needs.

moneyweb
dot arrow

By the year 2020, almost 60% of firms anticipate increasing their AI hardware/ software budgets by nearly 50%.

Workstation capabilities provide critical benefits to AI strategies offering key value in performance and security which firms feel are worth additional budget.

AI opportunities utilizing workstations are widespread across both large enterprises and smaller firms when data is stored locally

76% of firms have identified at least one scenario for AI in the last year and another 20% are looking for an opportunity.

magnified business
citystrip

These AI scenarios provide firms an opportunity to integrate workstations into their workflow; increasing performance while lowering costs and simplifying overall workflow.

Prioritize security, then assess your use cases to balance between speed and cost

Firms listed the following as the most important features for an AI platform:

  • 52% Performance
  • 50% Security
  • 42% Cost
  • They were rated more important than ease of use (40%), scalability (30%), and trained staff/manpower (28%)

dot arrow

Assess applications such as speech analysis or intelligent recommendations that can be run at a lower cost on a workstation over a long timeframe than on a cloud or server platform.

Workstations solve big challenges for firms

business people

Servers and cloud platforms have a longer history of use, but one third of respondents agree that workstations are the most effective platform on which to develop AI applications. This frees up highly-scaled servers and cloud platforms to be used for other initiatives.

Firms seek a platform for AI solutions that increases efficiency and productivity

chart

Over a third of firms (34%) rank efficiency and productivity as the most valuable strategic benefits to running AI applications on a platform that best fits their company's needs.

workstation
workstation

Workstations simplify workflows with ease of use and scalability creating both efficiency and productivity.


Workstations are essential as a development platform:

DEVELOPING AI APPLICATIONS

Base: 210 AI & Workstation decision-makers in North American companies with 100+ employees
Source: A commissioned study conducted by Forrester Consulting on behalf of Dell, April 2018


Many firms run core AI applications on workstations today

Workstations provide firms cost savings to train vertical AI solutions and provide excellent performance in a secure, local development environment making core apps like speech and image analysis fast and safe.

On which platform are you currently using to run core AI applications on at your organization today?

Base: 210 AI & Workstation decision-makers in North American companies with 100+ employees
Source: A commissioned study conducted by Forrester Consulting on behalf of Dell, April 2018


Firms need solutions that enhance internal skillsets

Regardless of what AI platform firms are using (servers, cloud, workstations, etc.), they listed a lack of skilled staff/training and lack of easy use amongst their top challenges with their current AI platform. Those challenges leave internal gaps. Companies seek access to technology round table discussions, containers with ready-to-go software, and services for hire.

*Containers are technologies that allow you to package and isolate applications with their entire runtime environment. Containers execute in a container runtime environment, and generally share an operating system.

48% access to technology round table discussions 46% Containers with ready-to-go software 29% Services for hire (training, installations, etc.)

Apply workstation capabilities to your industry's AI scenarios

For this study we focused on six industries’ specific AI scenarios, however many of those scenarios are applicable across other industries

Please describe the specific business scenario(s) your organization has applied AI to in the last year?

word cloud

Base: 210 AI & Workstation decision-makers in North American companies with 100+ employees
Source: A commissioned study conducted by Forrester Consulting on behalf of Dell, April 2018

Emerging AI technology opportunities on workstations are key for the Energy industry

Current Use:

Future Use:

Current Use:

Key energy use cases for AI on workstations are:

  • 44% Oil production optimization
  • 35% Prediction of consumption demand

Business problems organizations are using AI to solve today on workstations:

24% Classifying items within a data set

24% Deploying trained inference models (proofs of concept or production)

18% Guiding interactions with customers for common tasks

Future Use:

However, there is are great opportunities to explore emerging technology opportunities on workstations like:

38% Weather prediction

35% Wind power generation

24% Drilling and exploration

18% Solar forecasts

Manufacturing & Engineering keep factories safe and automated with AI on workstations

Current Use:

Future Use:

Current Use:

Key manufacturing and engineering use cases for AI on workstations are:

  • 50% Safety monitoring/analysis
  • 39% Factory and demand analytics

Business problems organizations are using AI to solve today on workstations:

32% Deploying trained inference models (proofs of concept or production

32% Developing and training AI algorithms to data

32% Experimenting and knowledge building

32% Recommending appropriate objectives to customers

Future Use:

However, there is are great opportunities to explore emerging technology opportunities on workstations like:

45% Smart manufacturing systems

24% Automated/generative design

18% Relationship intelligence

Content is the key AI use case for the Media and Entertainment industry on workstations

Current Use:

Future Use:

Current Use:

Key media and entertainment use cases for AI on workstations are content focused:

  • 50% Content-based search
  • 47% Content suggestions based on selections over time

Business problems organizations are using AI to solve today on workstations:

38% Recommending appropriate objectives to customers

28% Predicting trends or other variables

19% Developing and training AI algorithms to data

Future Use:

However, there is are great opportunities to explore emerging technology opportunities on workstations like:

47% Real-time translations

44% Language processing

41% Video captioning

Insurance and Financial Services rely on AI for targeting and IT

Current Use:

Future Use:

Current Use:

Key use cases for AI on workstations amongst insurance and financial services respondents are:

  • 58% IT management
  • 33% Operations

Business problems organizations are using AI to solve today on workstations:

31% Developing and training AI algorithms to data

22% Mining for customer sentiment

17% Predicting trends or other variables

Future Use:

However, there is are great opportunities to explore emerging technology opportunities on workstations like:

28% Sustainability and resource management

25% Risk management

6% Research and design

Emerging AI technology opportunities help the Hospitality & Travel industry win customers

Current Use:

Future Use:

Current Use:

Key hospitality and travel use cases for AI on workstations are:

  • 44% Customer service
  • 35% Marketing AI applications
  • 35% Price forecasting/pricing promotions

Business problems organizations are using AI to solve today on workstations:

35% Mining for customer sentiment

26% Recommending appropriate objectives to customers

26% Classifying item within a data set

Future Use:

However, there is are great opportunities to explore emerging technology opportunities on workstations like:

29% Sustainability and resource management

29% Risk management

29% Research and design

AI emerging tech opportunities drive Healthcare to new heights of diagnosis and innovation

Current Use:

Future Use:

Current Use:

Key healthcare use cases for AI on workstations are:

  • 44% Healthcare predictive system maintenance
  • 36% Financial analysis

Business problems organizations are using AI to solve today on workstations:

19% Developing and training AI algorithms to data

19% Detecting anomalies within a data set

17% Guiding interactions with customers for common tasks

Future Use:

However, there is are great opportunities to explore emerging technology opportunities on workstations like:

28% Diagnosing disease from patient scans

22% Cancer detection

14% Therapeutic treatment for stroke, spinal injuries, or prosthetics


Base: 34 AI & Workstation decision-makers in North American companies with 100+ employees
Source: A commissioned study conducted by Forrester Consulting on behalf of Dell, April 2018

Key Recommendations

AI strategists imagine a future where artificial intelligence will support a complex and diverse business infrastructure with autonomy. This vision is a decade or more off for most firms. To prepare for this future, IT decision makers and AI practitioners should take a focused and pragmatic approach to AI in order to prepare for the future they imagine.

Start by focusing on solving specific business problems with well-defined scopes and measurable outcomes. These problems have clear business relevance, utilize data that organizations have or can easily collect, and leverage practical AI software tools and deployment strategies.

Focused AI will involve both local computing as well as cloud and data center. Evaluate your overall infrastructure to be sure it matches the practical applications you wish to run. Benchmark yourself against peers for best practices in choosing workstations, cloud providers, and the like.

Highly-scaled servers and cloud platforms are necessary to run applications that must be run at top speed and where cost is not a barrier. However, workstations provide excellent support for applications where data security is a priority but where timelines are more flexible or cost is a major consideration.

Your peers showed a clear preference for these decision variables when thinking about AI efforts. Why? Performance speaks to the speed at which focused AI can actually solve problems; most often, that means speedily. Security is more crucial than ever, particularly with data sets that involve customer or other sensitive data. To meet increased security requirement, firms might need to trade off the time it takes to train and run algorithms. And of course cost helps keep the business case in line with goals.

Firms designing an infrastructure to run AI applications must balance performance against cost while always keeping security in mind. You can think of this as a scale: Security keeps everything in balance, and you must make trade-offs between performance and cost

Nearly a third of firms today see developing AI applications, training or processing algorithms, and R&D ideation or iteration as key workloads for workstations. Developers and data scientists’ workflows benefit from the flexibility, security, and speed of workstations for creating focused models, even if these are later run as cloud or data center workloads. Today, you can evaluate or expand your use of workstations to achieve secure, cost managed solutions that meet the needs for many AI applications.

Today, about a quarter of workstation decision-makers already say they use workstations to run core AI applications in pre-trained vertical solutions, intelligent research solutions, image and video analysis, and speech analytics. For scenarios that require the scale of the cloud, however, workstations can still play a key role, taking on data management, algorithm training, and ideation activities that will later be scaled out to the cloud. Open source algorithms, in particular, can be run on workstations and deployed to the cloud once the AI applications are ready for launch. Workstations can therefore be used in combination with data center, and cloud-based deployments during the development and training phases. They may offer a better overall solution where time to train or run a problem is a smaller consideration than budget.