Electronic design automation (EDA) workloads: using IBM Cloud

A market sector known as Electronic design automation(EDA) is made up of hardware, software, and services that are used to help define, plan, design, implement, verify, and then manufacture semiconductor devices (or chips). Foundries or fabs that produce semiconductors are the main suppliers of this service.

EDA solutions are essential in three aspects even if they are not directly engaged in the production of chips:

• To make sure the semiconductor manufacturing process produces the necessary performance and density, EDA tools are employed in its design and validation.
• EDA tools are used to confirm that a design satisfies every criterion related to the manufacturing process. We call this area of study “design for manufacturability” (DFM).
• The need to track the chip’s performance from the point of manufacture to field deployment is becoming more and more pressing. The term “silicon lifecycle management” (SLM) describes this third use.

EDA HPC environments are constantly expanding in size because to increased competition, the need to get products to market more quickly, and the growing demands for computers to handle the more fidelity simulations and modeling workloads. Businesses are trying to figure out how to effectively leverage technologies like hybrid cloud, accelerators, and containerization to obtain a competitive computing advantage.

The integrated circuit design and Electronic design automation software

In order to shape and validate state-of-the-art semiconductor chips, optimize their manufacturing processes, and guarantee that improvements in performance and density are reliably and consistently realized, Electronic design automation (EDA) software is essential.

For startups and small enterprises looking to enter this industry, the costs of purchasing and maintaining the computer environments, tools, and IT know-how required to use Electronic design automation tools pose a substantial obstacle. In addition, these expenses continue to be a major worry for well-established companies that use EDA designs. A key component of success is meeting deadlines, and chip designers and manufacturers are under tremendous pressure to introduce new chip generations with improved density, dependability, and efficiency.

The following picture illustrates how this problem in integrated circuit (IC) design and fabrication may be seen as a triangle opportunity space:

Industry difficulties for Electronic design automation

Three essential resources are crucial to design potential in the field of electrical design automation (EDA):

1. Compute infrastructure
2. Designers
3. EDA licenses

The most difficult task for design firms is deciding which projects have the best chance of fostering expansion and financial success. An expansion of the pool of designers, licenses, or compute infrastructure is necessary to maximize these potential.

1.Computing infrastructure

Extensive preparation and time are needed for the acquisition, installation, setup, and use of compute resources in order to grow the on-premises computing infrastructure. Bottlenecks in the compute market, the approval process for additional data center resources, and the building of infrastructure related to power, cooling, and electricity can all cause delays. Meeting the need for enlarged data centers fast requires outside help, even for major firms with sizable on-premises data centers.

2.Designers

The pool of designers is the second constraint that restricts realizable opportunities. Hiring designers who are highly qualified engineers quickly is a difficulty. Establishing the educational foundation necessary for design takes years, and successfully integrating new designers into established design teams sometimes takes a year or longer. This limits commercial options by making designers the most inelastic component on the left side of the figure.

3.Electronic design automation permits

Finally, contracts defining the permitted types and quantities of tools a business may employ typically regulate EDA licensing. Large businesses may be able to get almost limitless enterprise license agreements, while startups and small to medium-sized design companies cannot afford them.

Using cloud computing to accelerate the time to market

Cloud computing and new EDA licenses are two essential components that businesses of all sizes may strategically exploit to boost prospects and broaden their company horizons and achieve a competitive edge in terms of time-to-market.

Unlike internal infrastructure development, which takes months, cloud computing allows for the quick growth of compute capacity by providing or constructing new infrastructure in public clouds in a matter of minutes. Additionally, peak-licensing models are also being offered by EDA software providers, allowing design firms to use EDA software in the cloud for shorter periods of time than with typical licensing agreements.

By utilizing modern EDA licensing arrangements and cloud computing, most design companies may greatly increase the scope of their economic opportunities. Although organizations may improve their design efficiency by utilizing the automation benefits provided by EDA software and cloud computing infrastructure provisioning, the availability of designers is still an inelastic resource.

How EDA is being led by IBM

Together with IBM’s extensive knowledge of data, artificial intelligence (AI), and semiconductor technology, our wide range of EDA and HPC products includes systems, storage, AI, grid, and scalable task management. IBM data transfer and storage technologies, including IBM Storage Scale, IBM Spectrum LSF, and IBM Aspera, have won awards and are highly integrated to provide large-scale task management across various clusters and computational resources as well as high-performance parallel storage solutions.

With a single point of ownership, IBM Cloud EDA infrastructure provides foundry-secure environments and patterns. Secure, high-performance, and easily navigable cloud solutions built on top of IBM’s industry-leading cloud storage and task management infrastructure can immediately provide value to EDA organizations.

Technical leaders at IBM will release a white paper in the upcoming months that highlights IBM special capabilities to provide both big and small corporate clients with optimized IBM public cloud infrastructure for EDA applications.