Chip companies, once seen as engineering pure plays, are now at the center of geopolitical intrigue. Chip manufacturing firms, especially TSMC and Intel, have become the backbone of devices with an on/off switch.
These firms have also created a blueprint for product planning, development, and flawless execution of engineering products. Semiconductor veterans are now carrying their expertise to industries outside the chip industry.
Raj Talluri, the CEO of battery firm Enovix, is applying his lessons from companies like Qualcomm and Micron to the highly competitive battery industry. The company makes silicon-based batteries and competes with companies developing new chemistries for products that include electric cars, mobile phones, laptops, and edge devices.
Device batteries are still based on old technology, and companies have struggled to bring products to market due to issues with production and cost.
Enovix’s battery chemistry is ready, and Talluri is applying his chip and memory industry experience to tackle some of those challenges, including finding willing takers for its technology. Enovix recently signed a development deal with a phone maker and reached a manufacturing milestone.
Talluri sat down with HPCwire to explain how lessons in silicon and memory can be applied to the battery industry.
HPCwire: How are you applying a Silicon Valley chip mentality to accelerate battery development?
Talluri: Getting a chip mentality really helps. The tricks that the chip guys have done for a long time in getting to know the results quickly have not been done in the battery industry as much. We’ve developed an innovation where, based on cycling a battery 50 times, 100 times, 200 times and looking at the internal composition changes, we can predict how it will perform out to 1000 cycles. That’s an innovation we have developed here that I’ve been pushing for quite a bit since I came.
HPCwire: What unique challenges does Enovix face compared to traditional chipmakers you’ve worked at, such as Qualcomm and Micron?
Talluri: There are a few distinct challenges in batteries compared to chips. One of the challenges is what I call time to production. By the time you give customer samples, the time they actually get to high volume production is between nine and 12 months minimum, which means the time to develop a battery is multiple years before that.
In chips, we found out that you can simulate the kind of stresses the chip would undergo in this life stage very quickly in a short time. There is no such burden in batteries, so you can’t predict a battery will go 1000 cycles without cycling it 1000 times, which was weird when I first came into the market.
HPCwire: How is vertical integration of battery manufacturing similar to the memory chip business?
Talluri: If you look at the memory business, you don’t share factories; every memory maker has its own factory. It’s a true vertical integration. You have to understand the end customer’s needs, define a product, design the product, and manufacture the product uniquely that only you do, and others don’t. That’s been a little bit of how I feel here, and that’s why I feel so much more comfortable doing this after spending five years at Micron.
HPCwire: As a small company, how will you scale manufacturing to meet customer demand?
Talluri: Most of our competition is graphite cells with a little silicon powder. In that sense, we’re a little bit unique – nobody in the market today can make 100% active silicon anode batteries. That’s how we contest versus competition. That said, we have to build the factory because we have chosen that we don’t want to just license our technology, we want to build the factory. Fortunately, we have a lot of cash the company has raised over the years.
We’ll build a small factory in Fremont to produce some samples, but we are building our big factory in Malaysia. The first line can produce around 10 million batteries. We’re focused on proving out the technology and getting it shipping. I think there is enough capital in the market that once we do that, through partnerships, licensing, and other ways, we can add scale and capacity to meet customer demand.
HPCwire: Chipmakers rely heavily on electronic design automation tools and standard manufacturing processes. How are you dealing with the lack of a similar ecosystem in batteries?
Talluri: In chips, you have these tools from companies like Cadence and Synopsys for designing a chip, followed by standard machines from Applied Materials and Lam Research that build the chips. In batteries, those don’t exist. We have to do a lot of design experiments to choose the chemicals, anodes, cathodes, and electrolytes, put them together, and see how they work. Then, we build the machine that makes the batteries. In this ecosystem, we have to do all of it ourselves. That’s why batteries take so long and why they’re so differentiated.
HPCwire: Are you concerned about competing with well-established battery giants?
Talluri: When you look at the battery ecosystem, some people make the materials that go into the battery, people who make the cells, and people who make the packs. We are a materials-agnostic company. If there’s innovation in anodes, cathodes, and separators, we can use that. We’ve figured out how to combine different anodes, cathodes, and electrolytes to create the best-performing battery for the right application while reducing the silicon swelling problem with our unique battery architecture. Honestly, there is nobody in the market today that can make 100% active silicon anode batteries.
HPCwire: From your prior roles, you have deep relationships with consumer electronics companies. Is that helping Enovix gain traction?
Talluri: I have 30 years in this industry. I know most of the people who run this business, so I can call them directly, pick up the phone, talk to them, or visit them. The access has not been a problem. But the battery choice, especially for a phone, is a huge decision because it’s a small space that needs high energy density and safety. It has to meet all these requirements and last for years. It’s not a commodity; it’s a very high-value custom product. We have to work with them to make it exactly how they want.
HPCwire: Cost is a big barrier to battery adoption. How do you tackle that?
Talluri: Our cost is a little bit higher, but the energy density we provide is so much more that our price is higher, and people are pretty comfortable with that. If you make a $1000 phone, the bill of materials is typically $600. The battery is probably $6-8 today. But we can charge $10-12. For them, $10-12 for a $600 phone is not that much. But the improvement it provides in making a faster processor, faster memory… running more AI applications is huge. We unlock the value of other components in the phone. People don’t mind paying more for a battery, which unlocks the value of the rest of the ecosystem in the product.