It was only a decade ago when inventor Shay Benisty was first challenged to pursue a patent. Two, in fact. His manager added a new task to his annual performance goals: submit two patents each year.
Benisty, now a distinguished engineer leading Western Digital’s ASIC (Application-Specific Integrated Circuit) system architecture team, was intimidated at the time. He procrastinated, putting it off until the 11th month.
“It seemed like an unachievable goal. I didn’t know where to start, what the process was, and who could help me,” recalls Benisty.
Unsure of what to expect, he prepared 10 slides for his first internal review board meeting but later learned he’d only have a few minutes to present his idea. He boiled it down to its key advantage: it was a technology for optimizing random read commands in flash. It was a clever idea and had practical virtues. He not only aced the presentation, but his patent was later accepted. He knew then he wanted to pursue more.
Today, Benisty is a prolific inventor, holding more than 150 patents (and counting) and has become an expert in identifying and implementing patent-worthy innovations. In his office, he sits in a sea of hexagon-shaped patent award plaques displayed on his wall.
“Patents are important to a technology company,” said Benisty. “Not only because they protect intellectual property, but they also motivate and inspire inventors, and they drive innovations to the next level, enabling advancements in technologies and products.”
A passion for problem solving
Benisty hails from Beer Sheva, Israel, where he graduated Summa Cum Laude from the Ben Gurion University of the Negev with bachelor’s degrees in electrical and computer engineering.
As a youth, he had an affinity for engineering, although he was first in his family to pursue the field.
“Growing up it seemed obvious that I’d have a path in engineering because I had a deep love for math, physics, and computers,” said Benisty.
He recalls a school field trip where his teacher challenged the class to solve a mathematical riddle known as “the merchant and the fake coin.” The task was to figure out which of 13 coins were fake (not made of real gold).
“We could only use three weights to figure it out,” Benisty recalls. He pondered the problem for hours on the long bus ride across the desert. Ultimately, after his thorough analysis and creative thinking, he was the one student who solved it and won a prize.
Solving challenges with inventions
Benisty brings his passion for solving problems to his job each day.
At the helm of the ASIC system architecture team, he manages a global team of 20 engineers who together are responsible for the architecture of the company’s flash-based controllers. These chips manage the connections between the storage devices and their various interfaces, such as the host, memory, and the logic within.
“Some call ASICs the brains of storage, after the firmware sends the plan for what needs to happen,” said Benisty. “ASICs decide how to prioritize and optimize how the plan is executed, like an air traffic controller who directs pilots on the best path to take, at what speed, and how to approach the runway.”
In addition to enhancing controllers with new features, his team has spent the past few years unifying elements of the company’s different flash-based controllers’ architectures, from consumer to enterprise products.
“The idea is to have a single core architecture that can be easily adapted to all of the product lines, which brings huge benefits in terms of efficiencies and time-to-market,” said Benisty.
On the lookout for patent candidates
Since that first patent, Benisty has become an expert on what’s required for patenting an innovation, such as the thresholds that must be met and how to master the application process.
Most of his patents fall under the categories of performance and power usage, predictive capabilities, and improving system features for ASICs. Sometimes they combine all three.
One such example is his invention of “activity-based device-initiated state transitions.” The feature is designed to detect when a device (such as a laptop or a phone) will go idle and predict how much time it will need to “wake up.” A device armed with this information results in
a faster response time when exiting sleep state, delivering a higher quality of service to the end user. The technology highlights Benisty’s forte, finding the right trade-off between responsiveness and power consumption.
“Client devices may have internal or external DRAM and are more sensitive to power usage since they can be unplugged,” said Benisty. “On the enterprise side, where SSDs are housed in large data centers, the top concerns pertain to performance and quality of service.”
Bensity and team must continuously evaluate potential trade-offs and optimize a device’s architecture for both power and performance.
From idea to invention to patent
When asked about his approach to invention, Benisty said he thinks first about the problem it addresses and how to solve it.
“I believe a good patent candidate should be novel, and it should represent a significant advancement from the existing solutions,” said Benisty. “It should have a clear problem statement,” he adds, “and a patent that addresses a wider range of challenges can be even more valuable than one that addresses a small problem.”
Once the problem is identified, Benisty looks at several alternatives and makes a pro/con list. He considers any constraints, such as resources or the schedule required to hit milestones, and the best path to completion.
“I like to think about several solutions for a problem before deciding what to do,” he said. “Even if I’ve thought of one solution, it’s not enough.” It’s a philosophy he also applies to other areas of his life.
Benisty explained that when a team of engineers implements a new technology, there will be little time to file patents because they must stay on schedule to deliver new features. However, once they reach a milestone, Benisty encourages a proverbial “day off” to consider patent candidates.
“I tell my team to take a breath and think about what they’ve just done, what new technologies have been implemented, and what new idea could be patented,” said Benisty. “Reflect on those and then go file for patents!”
He also recommends creating working groups to brainstorm ideas and discuss solutions. And, when a list of ideas is generated, bring them to patent harvesting meetings to help advance those ideas.
“With collaboration we can unlock new ideas that open new directions and innovation,” said Benisty, recognizing that many of his patents are the results of joining forces with other teams (such as firmware or algorithms) who each provide their point-of-view to a solution.
“But what inspires me most is coming up with the creative ideas that are implemented into our products and give a competitive advantage,” he said.
Artwork by Chris Connolly