Ganesh Moorthy, CEO of Microchip Technology (Image: Microchip)
By Bolaji Ojo
What’s at stake:
The electronics industry is just emerging from another one of its infamous downturns, this one aptly dubbed a “super cycle” by Ganesh Moorthy, CEO of Microchip Technology. Is the high-tech sector destined to repeating this boom-and-bust cycles, or are there actions that can be taken to mitigate the severity of the swings? Moorthy talked about this at the recently held Business of Semiconductor Summit (BOSS 2024), analyzing the genesis of the latest downturn and offering suggestions for the future.
The edges of the next electronics cycle are already being sketched. We just don’t know it yet, says Ganesh Moorthy, CEO of Microchip Technology Inc.
This means the electronics supply chain must be creative in its response to upcycles and downcycles, preparing for the next demand and sales fluctuations with actions taken in the middle – rather than at the end or beginning – of an ongoing cycle.
“You must think through a cycle,” Moorthy told the Ojo-Yoshida Report. “If I were to use a sports metaphor and ice hockey as an example, you need to skate to where the puck is going, rather than where the puck is.”
In a discussion last month at the annual Business of Semiconductor Summit (BOSS 2024), Moorthy reviewed the current state of the electronics industry, conducting a forensic examination of the market’s latest cyclical swing and the implications of these developments.
The industry is prone to “whiplash,” Moorthy said, noting that cycles occur because of actions related to inventory and manufacturing capacity mishaps as well as the impact of economic, geopolitical and sometimes natural factors over which enterprises in the sector have limited control.
Today, the semiconductor market is on a recovery path after the latest downturn, which resulted in severe pricing erosions and double-digit sales decline at many chipmakers. The industry is still dealing with the excess inventories piled up by vendors during the post-Covid 19 boom. That upturn, fueled by manufacturers’ fears of not being able to satisfy roaring orders, resulted in double-orderings, stretched out lead-times and a race to add new production facilities that now – in hindsight – seems unnecessary and rushed.
“What distorted the demand signal for everybody was when lead times became very long,” Moorthy said. “We typically run under eight weeks but then it went to 52 weeks and sometimes longer than that.”
So, what should companies do when the typical upturn or downturn rolls in? Moorthy says executives should ignore factors beyond their control and instead focus on things they can do to accelerate efficiencies in their operations and assist customers and suppliers.
“You can’t always control lead times,” he said. “What you must focus on in a downcycle is what is controllable. ‘Will the Fed change interest rates; Will inflation be x or y and; Will China become an issue’” These are not controllable? We focus on the things we think are controllable, including: what do we do with our new product introductions, and how do we maintain innovation?”
Here are the key points Moorthy made in his discussion with Junko Yoshida, editor-in-chief, and Bolaji Ojo, managing editor:
Outline of the cycle
This is a cycle whose magnitude and length has been substantially higher and longer than the normal semiconductor cycle. To put it in perspective, Microchip experienced two consecutive years of over 25 or close to 25% growth, cumulatively almost 55% over two years. Now, we’re on the other side, where the depth and length of the downcycle is also more than what we have seen in any previous cycle. So, on both sides of the cycle, we saw much higher highs, much lower lows, and it’s taking much longer time. That’s why it’s a super cycle.
Assessing the roots
It’s been very painful on both sides [of the cycle] because the industry plans for a certain growth level. If you over plan, you’ll end up with excess capacity. If you under plan, you’ll have unfulfilled demand.
What we had was a bullwhip effect of everybody assuming that they needed more capacity from the supply chain. And that’s what drove those 12 quarters or so of substantial growth. As we have seen in other cycles, bullwhip effect, when it unwinds, can be quite painful because it distorts the demand.
In this case, demand was significantly distorted from consumption and now we’re on the unwinding part over the last four or five quarters. We are going through the hangover of what was bought but not consumed in that cycle.
Cycle psychology
The psychology of the shortages played out. People were building products that had high value, [and] they didn’t want a semiconductor component that was maybe less than 1% of the value of their BOM (bill of materials) to limit these high value products.
So, they bought more than they needed or placed orders with multiple sources, thinking that not everybody would come through. Because supply was constrained, they couldn’t get everything they wanted. There comes a tipping point when supply begins to catch up, where you receive more than you had expected, and before you know it, you’ve got a flood.
Visibility and solutions
Each cycle is different. You try to apply the lessons from the prior cycles, but new things come about in every cycle. It’s a capital-intensive industry, so we need to have some assumption of growth from which we make our investments, whether as an IDM or a foundry that is providing services to anyone else.
Everybody has an expensive decision to make, and you must pick a [growth] number that is reasonable based on your customer forecast. We’re never going to be precise, and you need to create a capacity which has a little bit of a buffer. It’s when something that is substantially outside of normal happens that it creates a discontinuity.
There will come a point when visibility gets better. Today, visibility is very poor. Orders will start to increase because the orders must get to back to consumption and that starts the cycle over again and again. We’ll attempt to determine the real consumption.
AI to the Rescue?
We need a way by which we can be more perceptive about the demand signals we’re getting. Now, there are better tools that we can we use, like AI, which can help because discerning demand signals is a pattern recognition problem.
How would we apply that? If we could go back to orders that we see as abnormal, from multiple places, we can use tools to better discern signals that just don’t make sense, and then confront customers with that data.
Partnerships as a solution
Where possible we can co-invest with some customers and suppliers in taking risk. Building capacity that is above a normal trend line is an expensive proposition. One of the tests we put people through is to say, ‘if you want the supply chain to build additional capacity are you willing to stand behind them’? We tried that and it has been successful with those players who can see where their longer-term demand is going to be. We begin to share risks and shared rewards for where we think demand is going to be.
Strategies for mitigating the impacts of downcycles
What you must focus on in a downcycle is what is controllable. ‘Will the Fed change interest rates; Will inflation be x or y and; Will China become an issue’” These are not controllable? We focus on the things we think are controllable, including what we do with our new product introductions, and how we maintain innovation.
What our customers need when they go through cycles is the ability to take advantage of innovations that allow them to reduce cost, bring out new capability and go into new markets.
The second thing we look at is what we are doing to help customers and ourselves convert activity into results. We help them get innovations qualified and create the opportunity to grow demand.
The final thing is how we are getting our internal capabilities into fighting fit shape; what we are doing on improvements in the metrics of efficiency and metrics of effectiveness in our factory, R&D, and in other areas.
We do need to take some of that energy and apply it to improving and sharpening the saw for the next cycle.
Best practices
There needs to be an element of counter cyclicality in management team’s thought process. You must think through a cycle. You can’t be rushing to do things for the upcycle during the upcycle and rushing to do things for the downcycle during the down cycle. In the middle of an upcycle think about what you must do to prepare for the down cycle. And, in the middle of a downcycle, prepare what you need in the upcycle. If I were to use a sports metaphor and ice hockey as an example, you need to skate to where the puck is going, rather than where the puck is.
Take solace in the fact that cycles end. As painful as it felt when you were being yelled at by customers and you couldn’t tell when you were ever going to be able to solve shortages, it ended. When you think: ‘My God, when are we going to get out of this hole?’ Have some historical perspective and take solace in the fact that it will end.
Bottom line:
Semiconductor cycles have a shelf life. They end, sometimes sooner than expected and, oftentimes later than anticipated. Preparations for future cycles must start earlier than the industry is used to doing nowadays, which involves a firefighting approach where defensive actions occur much later than needed. Microchip Technology CEO Ganesh Moorthy’s counsel is to stay alert, agile and focus on innovations customers will always require no matter the cycle.
Related articles:
The Electronics Supply Chain is Still a Hot Mess
Bolaji Ojo is publisher and managing editor of the Ojo-Yoshida Report. He can be reached at [email protected].
Copyright permission/reprint service of a full Ojo-Yoshida Report story is available for promotional use on your website, marketing materials and social media promotions. Please send us an email at [email protected] for details.