Cloud: Key to Driving Innovation

Article By : Syed Alam and Greg Douglass, Accenture

While the benefits of cloud are significant, many semiconductor companies are still lagging behind in adoption.

At a time where the semiconductor industry is struggling to address rising chip demand and supply constraints, it has become an imperative for semiconductor businesses to turn to cloud to accelerate innovation, ramp up productivity and gain competitive advantage.

Semiconductor companies have a dual relationship with the cloud industry as they are both the makers of the chips that power it and users of the cloud themselves. While the benefits of cloud are significant, many semiconductor companies are still lagging behind in adoption.

ADVERTISING

Up to now, most businesses have focused their cloud efforts on less sensitive, non-critical back office functions. However, migrating and transforming core, compute-intensive R&D and manufacturing applications to the cloud could help semiconductor businesses significantly boost productivity.

Where to start

While the benefits of cloud span every sector, every industry is different. Semiconductor businesses looking to further advance their cloud journey are bound to gain value in three key areas: electronic design automation, which is essential to chip development, manufacturing analytics, which will greatly impact productivity, and talent acquisition, which is another, less-talked about shortage the semiconductor industry has to address.

Put together, these three areas can significantly assist semiconductor businesses in addressing ongoing shortages and catch up to the ever-growing demand for chips across all industries.

Electronic design automation

Electronic design automation (EDA) applications are essential components of chip development, but they are highly resource intensive. The process is complex and involves dozens of applications. Different steps, from chip design to packaging to board-level design, each have their own challenges to cloud deployment and their own compute needs.

EDA applications can benefit significantly by being moved to the cloud, including:

Scalability and flexibility

Compute demand in chip design varies both cyclically and unpredictably throughout the design process. With automated provisioning and straightforward user interfaces, a team embarking on physical verification, for example, can quickly activate as many machines as they need. When the step is over, they return the resources to the cloud and the charges stop.

Out-of-the-box tools

Cloud service providers (CSPs) offer sophisticated tools and services that are ready for use out of the box in which engineers can simply use the prebuilt tools and concentrate on their design tasks. Cloud-based data hubs help unlock the enterprise-wide data and provide engineers with self-service capabilities to run real-time analytics, as well as more complex AI and ML models. Data curated and prepared once can be used many times across the enterprise to improve innovation.

Improved productivity and efficiency

Instant access to compute and storage resources allows engineers to become more productive. Capacity on demand enables simultaneous testing, which can boost efficiency significantly.

(Click on the image for a larger view).

Visibility and predictability

The cloud infrastructure provides the tools and visibility to build intelligence into systems. This goes beyond the dashboards and spreadsheets. Cloud tools, including AI and ML support intensive process modeling across multiple variables. These capabilities can be used to determine whether a project can be delivered on time or late — or what would be required to finish early.

Security

Chip makers data security concerns have been barriers to cloud adoption. However, today’s cloud solutions offer enhanced security and automation technologies not available for on-premise systems. Resources expended on security by the average hyperscale CSP far exceeds that of any one enterprise. Hyperscalers offer a comprehensive suite of automated tools, state-of-the-art best practices like encryption of data in flight and at rest.

Cloud analytics & manufacturing

The cloud is better equipped to handle the large amounts of data and data compute, from multiple sites and sources, needed to run analytics and AI around interpreting and connecting manufacturing fabs, as well as test data to create relevant product and engineering insights for improving product development and manufacturing quality.  Effectively combining data from multiple sites within an organization (both internal and external), as well as standardizing and normalizing for process, product and equipment differences, is only possible with the cloud’s ability to store data from multiple sites.

 Talent retention

The semiconductor industry has struggled for decades to recruit and retain tech talent due to the competition of working with cutting-edge technologies at younger, household-name startups and IPO-ready darlings. By embracing the cloud, semiconductor companies can attract the newest class of engineers who want to work with the latest technologies from the major cloud providers and develop new, highly marketable skills.

The more business functions that move to the cloud, the greater the benefits. Each company’s journey will be different. It requires a thoughtful look at which parts of the organization will deliver faster, better business outcomes as well as short-term and long-term cost savings.

These are five steps we recommend companies follow to start the transformation.

  1. Define clear business objectives for your cloud migration
  2. Design a sourcing strategy to decide which processes to keep running on-premise and which to migrate
  3. Assess your current applications for cloud readiness and decide which to re-platform, which to discard and which to migrate
  4. Develop an operating model that encourages the use of analytics and new ways of working.
  5. Create and execute a talent upskilling plan for your business and IT employees to understand the new capabilities that cloud-powered applications will enable

Semiconductor companies are in a unique position today to harness the power of the cloud for operational efficiencies and competitive advantages. Keep in mind, however, that any cloud transformation will require change across the organization. In addition to the above steps and strategies, getting the entire organization engaged, involved and excited about the migration will also be crucial to its success.

This article was originally published on EE Times.

Syed Alam is global semiconductor lead at Accenture.

Greg Douglass is Accenture’s global technology strategy and advisory lead.

Subscribe to Newsletter

Leave a comment