Who won the race?

Isn’t it easy to declare winners? Every race has a start and a finish line, and the one who finishes first is the winner. However, as competitions intensify, and athletes surpass boundaries, the differences among athletes are negligible. For instance, in the 2008 Olympics in Beijing, Michael Phelps won the 100m butterfly race by just 4.7 millimeters.

We need high-precision electronic devices to determine race winners accurately. Today we have electronic equipment that can measure in microseconds. In addition to measuring time, electronic devices are used in many arenas and fields, such as cars, high-performance computers, defense systems, AI applications, 5G phones, video-game systems, TVs, computers, and many more.

As electronic devices became pervasive in our lives, there was an increasing demand to enhance richer features with smaller and more elegant form factors. These devices are mostly manufactured by semiconductor companies that use electronic device automation (EDA) tools.

Electronic device automation tools are software tools used by semiconductor companies to design, simulate, verify, implement, and manufacture electronic equipment. Using EDA tools, semiconductor companies manufacture high-power devices such as Intel’s Itanium processor, with 1 billion transistors, as well as low-power devices such as tablets, digital watches, and many more.

There are many EDA tools that work during different stages of manufacturing – initial design, architectural simulation, logical simulation, timing, layout, mask generation, and fabrication. These tools generate thousands of design data and log files that require millions of CPU hours to complete the manufacturing process.

Here are some of the workloads that are commonly generated:

• Simulators that need extensive mathematical operations
• Transfer of large design databases from the customer to the EDA tools, and modified data transfers
• Data mirroring capability for large multisite design teams
• Sharing license files to save costs
• Production test on large volumes of data to verify defective chips

EDA workloads start with heavy computational jobs and end with data-intensive workloads. As a result, compute, network, and storage become important components of the infrastructure. Also, due to the wide range of products manufactured it is impossible to go with a one-size-fits-all approach. Flexibility, cost, and performance of the compute, storage, and network are essential for deploying EDA workloads.

NetApp has partnered with Equinix to provide a hybrid cloud architecture for electronic device manufacturers.

This architecture—NetApp^® Keystone^® Flex Subscription—deployed and monitored by NetApp and powered by Equinix, offers the following benefits:

• Equinix bare-metal servers deliver hardware compute for better performance.
• NetApp Keystone service deploys different types of storage—block and file, with the ability to increase or decrease based on demand.
• Equinix fabric connects public clouds (if needed), compute, and storage with low latency.

The EDA industry is facing unprecedented demand to manufacture different types of complex devices. The COVID-19 pandemic has caused a shortage of chips for most industries. As a result, this industry is ramping up production and in need of an infrastructure that is flexible, scalable, and reliable. NetApp Keystone Flex Subscription, along with Equinix, fills this need by providing the necessary infrastructure.

To learn more, visit the Flex Subscription page.

Read these blogs:

• NetApp delivers storage as a service, powered by Equinix
• NetApp Keystone at Equinix Delivers Storage-as-a-Service (Equinix blog)
• A car without a steering wheel
• Where is my elixir