The numbers Brief Episode map Takeaways Charts Quotes Clips Glossary Go next Transcript

Episode 102 · Deep tech · 37 min

The technologies that friction sets free

TDK Ventures invests at the atom layer — material science, not software — and its president argues that a fragmenting world is quietly the best thing to happen to clean technology: every tariff and chokepoint hands a Challenger technology like sodium-ion storage or metals recycling the shot at profitability it never had. The fund went into AutoFlight in 2020 when air taxis sounded absurd, and into Ascend Elements when it was ten people. Last year it wrote its first three cheques in India.

Deep tech & hardware Climate & energy Venture capital

Nicolas Sauvage

President, TDK Ventures · with Vishal Krishna

36:37

Said in this episode

▶ 5:31

88 years

TDK's age when this was recorded

A university spin-off funded by an angel investor that went B2B ferrite materials, then the cassette, then back to B2B batteries, sensors and components.

▶ 35:23

~7,000

Startup conversations on record

Every new TDK Ventures joiner gets access to nearly everything since the fund started in 2019 — decisions, improvements and conversations with close to seven thousand startups.

▶ 18:38

India investments in the past year

Sauvage's own verdict is that the fund should have come to India five years earlier; the sweet spot he names is seed to Series A.

▶ 8:08

10 people

Ascend Elements at the time of investment

Battery recycling had barely a market then; TDK's and ATL's in-house battery experts validated that the technology was credible.

▶ 17:19

2 professors

AM Batteries when the cheque went in

Two part-time professors, in Sauvage's telling — now, he says, one of the best solvent-free battery process companies around.

▶ 13:11

23%

Share of India's energy from renewables

Host Vishal Krishna's figure, offered while arguing India has yet to move to renewables at scale; not independently sourced on air.

The brief

The argument in sixty seconds

Sauvage's claim is that deep tech means atoms, not bits — innovation down at the material-science level, with hardware and software as two sides of one coin — and that it only earns its keep when it becomes profitable, because unprofitable technology cannot scale. From that first principle follows the contrarian move: back what is not yet obvious, roughly two years early, before consensus prices it in. AutoFlight went in during 2020 when air taxis sounded like science fiction; Ascend Elements was ten people; AM Batteries was two part-time professors. The sharper argument is about geopolitics. The fund deliberately models both a frictionless world and a fragmenting one, and Sauvage's conclusion is that friction acts as a subsidy: when rare materials get stuck behind tariffs, Challenger technologies — sodium-ion storage, copper and platinum-group recycling, green hydrogen, eventually nuclear fusion — finally get an opening against cheap incumbents. India sits in the middle of that. Its engineering brand power now stays home instead of emigrating, its temperature range makes sodium-ion interesting, and the global South's catch-up, which he concedes is unfair, is precisely where a profitable green technology could make India a powerhouse. What is missing is the exit — the hard-tech success story that would trigger the snowball SaaS already got. Three India investments in, an 88-year-old cassette company's venture arm is betting it arrives.

Worth your time if you are

Scientists sitting on a technology that isn't profitable yet

Deep-tech founders choosing between a grant and a term sheet

Climate and energy investors mapping post-lithium storage

Corporate strategy teams weighing an India venture arm

Episode map

Where the conversation travels

Every block is a chapter, coloured by what it's about. Click any of it to jump straight to that minute on YouTube.

Deep tech & hardwareClimate & energyVenture capitalIndia macroRegulation & policyLeadership & org

01Cold open: deep tech at the atom level 0:00 Vishal frames the episode around challenging technologies, a greener planet and corporate VCs, then Sauvage defines deep tech not as data and AI but as material science — atoms rather than bits, with hardware and software as two sides of the same coin. 02Microelectronics, sensors and the generation stress test 2:09 An engineering education in microelectronics before it was fashionable, a route into TDK through the MEMS-sensor company InvenSense, and the fund's habit of stress-testing whether a technology's materials survive not one or two but four generations of scale. 03Deep tech is unprofitable first, profitable eventually 4:10 Vishal recalls twenty years of Indian atom-level science living in academia and philanthropy; Sauvage's correction is that deep tech is rarely profitable to start — which is what corporate VCs are for — but must become profitable or it runs out of money before it scales. 04An 88-year-old cassette company's DNA 5:31 TDK began as a university spin-off funded by an angel investor, went from B2B ferrite materials to the cassette every Indian remembers and back to B2B batteries, sensors and components — which is why a venture arm helping entrepreneurs fits the parent's contribute-to-society DNA. 05Two years before it's obvious 6:37 Investing from first principles means finding what isn't obvious yet, because obvious is already priced in — hence AutoFlight in 2020 when people said air taxis were crazy, and Ascend Elements at ten people, when battery recycling had barely any market. 06Playing devil's advocate with geopolitics 9:13 The fund models both a frictionless globalised world and a fragmenting one, and finds that in the harsher scenario technologies with no chance against cheap incumbents suddenly get room to scale — the rare-materials chokepoint being the clearest slice. 07Challenger technologies, defined 12:04 Sauvage's key message to entrepreneurs: recycling of batteries, copper and platinum-group metals, nuclear fusion and sodium-ion storage are Challenger technologies — unprofitable today, but candidates to reach profitable scale precisely because of geopolitical friction. 08Fission is easy and unsafe; fusion is hard and safe 13:42 Wind and solar are intermittent, so storage complements them, but only nuclear supplies continuous power — and fission was doable decades ago yet inherently unsafe, while fusion demands the best minds and is inherently safe, the technology Sauvage thinks a society a century out would thank us for. 09Green hydrogen, green solvents, two professors 15:31 The green-hydrogen bet is to find the applications where it can be profitable first and displace fossil-fuel consumption, alongside PH7 Technologies recycling copper and platinum-group metals with a green solvent — and AM Batteries, backed when it was two part-time professors. 10Seed to Series A, and what goodness means 17:36 India today is not the India of five or ten years ago, three Indian investments went in last year, and the fund's sweet spot is seed to Series A because that is where it can shape the outcome — with 'TDK goodness' defined strictly from the recipient's point of view. 11Scaling is the part scientists have never done 19:40 Recruiting, picking investors who add value beyond capital, meeting customers and signing contracts is the hardest stretch of any venture — and taking VC money is a commitment to go fast and scale high, which Sauvage says is not for every entrepreneur. 12Partners, and the brand power that stays home 21:32 Venture capital reveals quickly who is genuinely good for founders, so the plan for the event is to pull more corporations and foreign corporate VCs into India — where the change is not the quality of engineers but the fact that their brand power no longer leaves. 13Brazil's 1975 ethanol bet, and India's version 23:29 Asked about Brazil choosing ethanol in 1975, Sauvage restates the Challenger thesis and offers India its own candidates — nuclear fusion and sodium-ion storage, the latter suited to the country's wide temperature range — with a message to scientists not to be discouraged by present unprofitability. 14The grid can't carry what's coming 26:03 Energy demand only scales and no country's grid can support it, between EVs, home electrification and the data centres feeding generative AI — which is also why the fund backs an inference chip company running large language models far faster than existing silicon. 15Beyond jugaad, and the exit that starts the snowball 28:00 Constraint-driven creativity is turning into creativity with hard tech; the global South's catch-up is unfair but is exactly where India could lead on green energy — and what hard tech still lacks is the SaaS-style exit that would pull in outside money and produce serial founders like the team behind Exponent Energy. 16A recent conversion and five dysfunctions 31:40 Sauvage says his climate conviction is recent and came from working with his team, then walks The Five Dysfunctions of a Team from vulnerability-based trust through conflict, commitment and accountability to results — the basis for radical transparency across nearly 7,000 startup conversations since 2019.

Takeaways

Ideas to carry out of this hour

Deep tech is atoms, and it only counts when it scales profitably

Sauvage refuses the Indian default that deep tech means data, AI and software. For TDK Ventures it is material science — innovation at the level of atoms — with hardware and software as two sides of one coin, which is why so much of the portfolio is climate tech. The corollary is unsentimental: most deep tech is not profitable at the start, which is exactly the gap corporate VCs and universities exist to bridge, but a technology that never reaches profitability cannot scale to the benefit it promised, because at some point you run out of money.

▶ 1:05 ▶ 1:38 ▶ 4:43

Be right two years before it's obvious, or pay for the consensus

The fund's stated method is first principles plus timing: if a thesis is already obvious to everyone, it is priced into the entry. AutoFlight went in during 2020, when most people said air taxis were never going to be needed; Ascend Elements was ten people with a recycling market that barely existed, validated on battery chemistry by TDK's and ATL's in-house experts. Both look obvious now — that lag is the return.

▶ 7:21 ▶ 7:36 ▶ 8:08

Geopolitical friction is a subsidy for Challenger technologies

TDK Ventures deliberately runs both scenarios — a frictionless globalised world and one with tariffs, wars and blocked material flows — and the second is where the thesis gets interesting. If rare materials route through a country that starts restricting them, the incumbent technology built on those materials becomes fragile, and technologies that had no chance on cost suddenly do: sodium-ion instead of lithium, recycled copper and platinum-group metals instead of mined ones. A Challenger technology is one that isn't profitable today but could be, and friction is its catalyst.

▶ 9:29 ▶ 11:33 ▶ 24:03

Fusion is hard because it is safe, and that is the point

Wind and solar are intermittent, storage complements them, but only nuclear gives continuous clean supply. Sauvage's framing inverts the usual difficulty argument: fission is not actually that difficult, which is why it was done decades ago, but it is inherently unsafe; fusion is extremely complex and demands the best minds, and is inherently safe. If a society a hundred years from now looks back gratefully at one investment, he thinks it is fusion — green, continuous, cheap energy that would itself rewrite a geopolitics currently organised around petroleum.

▶ 14:16 ▶ 14:31 ▶ 15:02

Scaling, not science, is where deep-tech founders break

The hardest part of any venture is scaling, and for scientists it is the part they have never practised. It means recruiting the right people, choosing investors who add value beyond capital while avoiding the bad ones, meeting customers and partners, signing contracts — and above all trusting others with a project you have thought about alone and deeply. Which is also why he warns that VC money is not for every entrepreneur: taking it is a declaration that you will go fast and scale high, and the thing stops being a project on the side.

▶ 19:56 ▶ 20:27 ▶ 21:00

India's engineering brand power finally stays in India

Nobody abroad doubts the quality of Indian engineers or the political will to build infrastructure — that was true before. What changed in the last five years, Sauvage argues, is that the brand power now stays home: the smart young minds who would have left create companies here instead, innovating for India with global ambition. He thinks most corporations have not absorbed the full implication yet, which is precisely why he is trying to drag more corporate VCs into the country, and why he says the fund should have arrived five years ago.

▶ 18:21 ▶ 18:38 ▶ 23:00

Hard tech is waiting for the exit that starts the snowball

Asked about selling to a strategic acquirer, Sauvage reframes it as India's missing ingredient. Investors crowd into Indian SaaS because SaaS produced success stories; hard tech has not yet produced its equivalent — an entrepreneur who started something genuinely difficult at the atoms level and exited well. That single event triggers the snowball: outside money follows, and the founder himself usually goes again at a harder problem, which is how you get the serial founders the ecosystem actually compounds on.

▶ 30:23 ▶ 30:40 ▶ 30:55

Goodness is scored by the recipient; trust is a team property

'TDK goodness' has a deliberately awkward definition — help given to an entrepreneur that the entrepreneur then confirms was valuable, measured from the recipient's side rather than the investor's deck. The same instinct runs the fund internally via The Five Dysfunctions of a Team, whose first rung Sauvage says is the most misunderstood: trust is not you-and-me reciprocity but whether you will expose your doubts and mistakes to the whole team. Only then can you disagree safely, commit, hold each other accountable and get results — enforced at TDK Ventures by giving every new joiner access to nearly everything since 2019.

▶ 19:09 ▶ 33:48 ▶ 35:07

The numbers, drawn

What the episode measures

Every figure below was said on air — timestamps included, caveats kept.

Conversation share

portion of the hour spent on each theme

Deep tech & hardware24%

Climate & energy20%

Venture capital18%

India macro14%

Regulation & policy8%

Leadership & org8%

Computed from the chapter map of this episode.

Worth keeping

Lines that stay

If it's obvious to everyone already, then it will be priced into our investment. You want to be right maybe two years before it's obvious to everyone.

— Nicolas Sauvage ▶ 7:21

At the end of the day every crisis is an opportunity — and as much as we hate what's happening in the world today, there is an opportunity in what that means for a cleaner planet.

— Nicolas Sauvage ▶ 11:41

Nuclear fission is actually not that difficult to do — that's why we could do it decades ago — but it's inherently unsafe. Nuclear fusion is the opposite: extremely complex, and inherently safe.

— Nicolas Sauvage ▶ 14:22

When you take VC money you're basically saying: I will go fast and I will scale high. It's no longer going to be a project on the side.

— Nicolas Sauvage ▶ 21:00

We're going beyond jugaad now. It's no longer about creativity with different constraints — it's creativity with technology, with hard tech, with different constraints.

— Nicolas Sauvage ▶ 28:35

Clips that travel

Short on time? Start here

Investors sizing early-stage deep-tech bets

Two years before it's obvious

AutoFlight in 2020 when air taxis sounded crazy, Ascend Elements at ten people — the clearest statement of how first-principles conviction beats consensus pricing.

6:37 → 9:13 · 3 min ▶ Watch clip

Founders working on technologies that aren't profitable yet

Why friction is a gift to Challenger technologies

The devil's-advocate globalisation exercise, the rare-materials chokepoint, and the definition of a Challenger technology — the argumentative core of the episode.

9:13 → 13:42 · 4 min ▶ Watch clip

Anyone weighing India's clean-energy options

Fission, fusion and the hundred-year thank-you

Intermittency, storage and the inverted safety argument: why the hard technology is the safe one, and what cheap continuous power does to geopolitics.

13:42 → 15:31 · 2 min ▶ Watch clip

Scientists thinking about leaving the lab

Seed to Series A, and what goodness means

Three India cheques, why the fund goes in early enough to shape outcomes, and the blunt warning that venture money is not for every founder.

17:36 → 21:32 · 4 min ▶ Watch clip

Indian hard-tech founders and their would-be backers

Beyond jugaad, and the exit that starts the snowball

Why the green-energy opening for the global South is real, and why one hard-tech exit would change the funding maths the way SaaS already did.

28:00 → 31:40 · 4 min ▶ Watch clip

Glossary

The jargon, unpacked

Deep tech

In Sauvage's definition, innovation at the level of atoms — material science — with hardware and the software on top of it treated as two sides of one coin, rather than the bits-only world of SaaS.

Corporate VC (CVC)

A venture fund run by an operating company rather than an independent partnership; TDK Ventures is TDK's, and its pitch is technical validation and go-to-market help from the parent alongside the cheque.

Challenger technology

A technology that is not profitable today but could become profitable if geopolitical friction reprices the incumbent it competes with — sodium-ion storage, metals recycling and nuclear fusion are the examples given.

TDK goodness

The fund's internal term for help given to a founder that the founder afterwards confirms was valuable — scored from the recipient's point of view, not the investor's.

Sodium-ion battery

An energy-storage chemistry that replaces scarce, geopolitically exposed lithium with abundant sodium; not yet profitable, but pitched here as well suited to India's wide temperature range.

eVTOL

Electric vertical take-off and landing aircraft — the 'air taxi' and air-cargo category TDK Ventures entered via AutoFlight in 2020, when the idea was widely dismissed.

PGM (platinum-group metals)

The scarce catalytic metals recovered by portfolio company PH7 Technologies using a green solvent; needed, among other things, for hydrogen fuel solutions.

The Five Dysfunctions of a Team

The management book Sauvage recommends: a ladder from vulnerability-based trust through conflict, commitment and accountability to results, and the template for TDK Ventures' radical transparency.

Connections

If this resonated, go here next

EP 114 · Deep tech

Full transcript

The whole conversation, searchable

Auto-generated captions, lightly cleaned. Click a timestamp to open that moment on YouTube.