That is an ongoing sequence on traders centered on rebuilding the bodily layer. The primary interview within the sequence was with Peter Barrett, a decade-long investor at Playground International.<\/p>\n
Mike Schroepfer based Gigascale Capital after departing Meta as CTO in 2022. The agency invests in corporations rebuilding the bodily financial system. As Schroepfer and his companions on the agency see it, surging demand for AI, energy and industrial capability is making a once-in-a-generation alternative to rebuild the bodily financial system \u2014 from vitality infrastructure and superior manufacturing to supplies and robotics. And as AI makes software program cheaper and simpler to create, the aggressive benefit more and more shifts to the {hardware}, vitality techniques and provide chains that underpin all of it.<\/p>\n Key to beginning the fund was Schroepfer\u2019s expertise constructing out the infrastructure to help Meta\u2019s enterprise. \u201cI might see the tendencies coming. We\u2019re going to want all of the compute,\u201d he mentioned. \u201cI don\u2019t know the place we\u2019re going to get the ability, so it\u2019s going to create this huge supply-demand crunch.\u201d<\/p>\n Gigascale raised its first institutional fund this month, a $250 million funding automobile. The agency has already made greater than 25 portfolio investments to this point.<\/p>\n Schroepfer\u2019s companions on the agency are Victoria Beasley, beforehand an investor at climate-focused investor Prelude Ventures, and Evaline Tsai, beforehand at Wonderful Construction Ventures.<\/p>\n Earlier than elevating the fund, the agency made 22 investments funded by Schroepfer\u2019s household workplace to be able to show the mannequin. On the time, the broad notion was you could possibly not make cash investing within the {hardware} layer.<\/p>\n Gigascale invests at pre-seed by means of Sequence A with some later-stage investments. Its examine measurement is anyplace from $1 million to $10 million.<\/p>\n {Hardware} companies will not be the identical as software program companies. \u201cIt\u2019s not a software program enterprise with increased capex,\u201d mentioned Schroepfer. \u201cThe failure modes are very totally different. The best way you intend and check and iterate, and what you perceive could be very totally different.\u201d<\/p>\n In our dialog, we spoke about an array of matters, together with vitality as a significant funding focus, his learnings from operating Meta, why now is a good time to construct a hard-tech firm and what excites him in regards to the SpaceX IPO.<\/p>\n Gen\u00e9 Teare: What’s Gigascale\u2019s thesis?<\/p>\n Mike Schroepfer: It\u2019s actually easy. We’re backing corporations which might be rebuilding the bodily financial system. That is how issues are powered, constructed, moved, manufactured and the way individuals are fed.<\/p>\n The assumption is there’s a confluence of technological modifications which might be bringing new merchandise and new corporations to market which might be higher, sooner, cheaper than what\u2019s on the market. That is the largest a part of the financial system.<\/p>\n One other solution to say it’s that we predict the long run is atoms, not bits, and it\u2019s a very thrilling time to be constructing these corporations.<\/p>\n What did you see that made you determine to arrange the fund in 2023?<\/p>\n Schroepfer: Plenty of the tech tendencies I’ve been a part of \u2014 from the net transition once I labored on Firefox, to the early internet infrastructure at Fb, to the cell transition within the early 2010s, to founding the Fb AI Analysis Lab in 2013, properly earlier than ChatGPT \u2014 have been wanting on the very shallow a part of exponential curves. These technological modifications didn’t appear that prevalent but, however they have been on this huge upswing.<\/p>\n I noticed the identical set of curves in photo voltaic cells, batteries and electrolyzers. They have been all going by means of huge exponential price downs, and on the identical time a large enhance in demand. We had electrical autos exhibiting up, onshoring and manufacturing, and this was pre-data facilities. I knew compute demand was going to develop. The place are we going to get the electrons to gas all of this? It\u2019s going to create an immense provide chain crunch.<\/p>\n Demand and provide have been converging on the identical time to create huge tailwinds. It simply felt like this chance to rebuild your complete bodily infrastructure in a means that our children are completely happy about. Which means, the brand new answer wins as a result of it’s cheaper, higher and sooner.<\/p>\n The opposite co-benefit it brings together with it’s that as a result of it’s easy and cheaper, it is usually much less polluting, so it doesn\u2019t damage people. I can construct a photo voltaic farm means sooner than I can construct a fuel energy plant. I can dwell subsequent to a photo voltaic farm and get zero air pollution. I don’t need to dwell subsequent to a fuel plant.<\/p>\n What I perceive in regards to the agency is that you’re very centered on vitality particularly. Is {that a} misunderstanding?<\/p>\n Schroepfer: It’s in all probability the only largest space that we spend money on. A big chunk of our portfolio is vitality. It’s a $2 trillion market and it’s the place the place I believe all of the disruption is occurring. However we additionally spend money on business, together with supplies from neodymium to copper, manufacturing and recycling, to quite a lot of AI within the bodily world. That features all the things from how I take advantage of AI to make my home extra environment friendly with HomeBoost, to how I construct power-efficient AI inference chips with Fractile.<\/p>\n Then there’s the constructed atmosphere, by way of buildings, and a bit of bit in meals. We do some bit in all the things, however in case you take a look at our portfolio, the 2 largest hunks are actually vitality and AI within the bodily world.<\/p>\n When do you assume Silicon Valley woke as much as the give attention to the bodily world?<\/p>\n Schroepfer: Within the broad consensus, it occurred just lately \u2014 within the final six to 12 months. There have been some of us who have been it early, however I believe the broad consensus has simply occurred just lately.<\/p>\n The opposite factor that I noticed is, if AI goes to make software program practically free to put in writing, then I believe software program companies is likely to be challenged, and the moat strikes to the {hardware}. The sport turns into: How do I get the infrastructure constructed to have a greater AI? That’s principally an infrastructure {hardware} downside, much less of a software program coding downside, and that’s going to filter by means of quite a lot of companies.<\/p>\n Once I began, frankly, three years in the past, I had many individuals \u2014 I’m pondering of somebody sitting in my workplace \u2014 saying, \u201cdon\u2019t do that.\u201d All the cash is in software program. You possibly can\u2019t make cash in {hardware}.<\/p>\n It doesn\u2019t damage that Cerebras, Fervo, SpaceX, Nvidia and TSMC at the moment are family names of corporations which have had huge valuation runs as a result of they’re such a core a part of the bodily financial system. I used to make use of Nvidia as my instance, however now I can use SpaceX. Speak about an organization within the largest market that’s operating away from the competitors. It\u2019s a very arduous firm to compete with.<\/p>\n How ought to we perceive the vitality wants within the U.S.?<\/p>\n Schroepfer: We\u2019ve been at comparatively flat demand over the previous 20 years or so, that means every year that goes by, we don\u2019t want way more energy, near 0%. We at the moment are rising at no less than a couple of % a yr.<\/p>\n One thing has gone from nearly no progress to comparatively excessive progress. You\u2019ve obtained tons of of gigawatts of information facilities deliberate to be constructed over the subsequent 5 years alone. That doesn\u2019t rely EV charging stations and electrification of properties and factories. It\u2019s a large supply-demand imbalance proper now, and constructing energy takes a very long time. When you\u2019ve obtained to construct an influence line, in case you\u2019ve obtained to allow a fuel energy plant, this stuff take years, not months. It has created huge demand, however everybody needs compute yesterday.<\/p>\n Meta has used tents as an alternative of buildings for his or her servers as a result of slicing out the time erecting metal for the constructing will get them compute sooner. Everybody is considering learn how to get energy sooner and learn how to get compute sooner as a result of, once more, it\u2019s a aggressive benefit when infrastructure is the moat.<\/p>\n Which applied sciences are you centered on within the shorter time period, after which the long term?<\/p>\n Schroepfer: We’ve got corporations deploying issues now. Within the energy crunch, one of many massive issues is that the demand for energy swings way more extensively than it used to. It was pretty regular. Now you will have massive coaching runs, you will have photo voltaic that comes on and comes off, and also you want a shock absorber to dampen the ability or cope with three or 4 days of clouds or no wind, in case you\u2019re relying on renewables.<\/p>\n Kind Vitality is an organization that has a brand new sort of battery that lasts for 4 days. You cost it up, and it\u2019s there for 100 hours. In any occasion the place an influence plant is offline or the solar will not be shining, Kind Vitality is there. Utilities consider this as an alternative of constructing a fuel energy plant. There are these fuel energy vegetation referred to as peakers, which you solely activate when you really want them. They sit there on a regular basis, and then you definitely fireplace them up in these intervals. As a substitute of doing that, which could be very costly, you will have this Kind Vitality battery: zero emissions, less expensive to function, and constructed from the bottom up for utilities utilizing a very totally different know-how. They will be deploying batteries this yr, for instance.<\/p>\n Entering into a distinct course, your complete provide chain for the way we get electrons to a constructing. I\u2019m going to construct a brand new information heart, and I’ve to hook it as much as the grid to get electrons there. There may be all this tools within the center referred to as energy transformers, these massive inexperienced packing containers or massive steel packing containers. It\u2019s actually Thirties know-how. We haven\u2019t modified a lot since then. They’re back-ordered for years now as a result of they’re these beautiful {hardware} machines.<\/p>\n There’s a new firm, Heron Energy, that mentioned, \u201cwait a second, we\u2019ve been delivery this new era of know-how referred to as solid-state energy electronics in electrical autos \u2014 the Mannequin 3, Mannequin Y, and extra \u2014 for hundreds of thousands of models a yr, with very quick ramps. We\u2019re taking that very same know-how and placing it on the grid.\u201d We\u2019re changing this Thirties know-how with 2020s know-how. It\u2019s extra environment friendly, it\u2019s a 3rd the dimensions and, most significantly, they\u2019re going to begin delivery plenty of models subsequent yr. They\u2019re constructing their manufacturing facility proper now. In 2027, they\u2019ll be delivery plenty of these Heron Hyperlink models.<\/p>\n Slightly bit additional out, we’ve got an organization referred to as Panthalassa that mentioned, \u201cwe\u2019ve obtained about 10 terawatts, which is an immense quantity of energy, within the Southern Ocean in waves sloshing round with nothing else happening down there. If we will harness that, it’s an untapped useful resource.\u201d<\/p>\n They\u2019re constructing autonomous buoys that float within the ocean. They bob up and down and switch that wave movement into electrical energy. Then they use that to energy, on the buoy, a compute node to do AI inference and use Starlink to ship the bits again. They\u2019re sort of exporting electrons through tokens within the Southern Ocean.<\/p>\n They\u2019ve been testing off the coast of Portland, they usually\u2019re going to deploy their first models subsequent yr. Folks have talked about information facilities in house. My massive pitch for this firm is that it\u2019s 100x cheaper to place a ton of capability within the open ocean than it’s to place it into house. When you assume information facilities in house are a good suggestion, you may need to take a look at the ocean.<\/p>\n Then you possibly can take into consideration Radiant, an organization in El Segundo, California. They’re constructing a compact, next-generation microreactor, or nuclear reactor. You possibly can consider it as one thing you placed on a truck or on an airplane, and it could actually run and energy one thing for 5 years straight. As a substitute of, in a distant area in Alaska or on a Pacific island, doing what they do now, which is delivery diesel gas there to run a diesel generator 24\/7, you put in considered one of these packing containers, and it produces energy for 5 years earlier than it wants refueling. Most significantly, once more, you wouldn’t need to sit subsequent to a diesel generator whereas it\u2019s working. It has very poisonous emissions. This factor has no emissions. It\u2019s good for people, and it\u2019s really going to be price aggressive with these issues. These are some examples of issues we\u2019re doing within the energy sector that I believe are actually affecting the long run.<\/p>\n Is there an unlock on this business that has made improvement cheaper and sooner at this second in time?<\/p>\n Schroepfer: The analog I\u2019d use is from computing. We used to construct mainframes, these massive building-sized computer systems. Then we had minicomputers that have been nonetheless actually massive. That is the motherboard for the primary server we designed at Meta that we deployed in 2011, referred to as Freedom. It was a Kind 1 server. It was the net server.<\/p>\n I put in hundreds of thousands of those, perhaps tens of hundreds of thousands. I don\u2019t even know what number of. They\u2019re all the identical, each single considered one of them. They go in a pizza-box-size factor that goes right into a rack in a constructing. That constructing is available in 4 models. Every of these is similar. That constructing is subsequent to a different constructing, which is strictly the identical. We construct 4 of these on a web site. All of them look the identical. I did that in 17 locations world wide. All of them look the identical.<\/p>\n The approach we use to make issues low-cost is mass manufacturing. Every little thing in your life that has gone down in value or improved in price-performance is mass manufactured: your iPhone, the servers and information facilities. They\u2019re all the identical. They\u2019re mass manufactured.<\/p>\n The world is filled with customized, bespoke stuff that\u2019s wickedly costly.<\/p>\n Within the energy grid, for instance, the entire stuff I talked about, you customized order it. I desire a transformer. I do engineering design. I ship it off to somebody. 4 years later, a truck reveals up with the crane and all the remainder of it. That\u2019s inherently costly and will get dearer yearly. Every little thing that’s customized will get dearer yearly, so I believe the largest factor we\u2019re seeing is that this transfer to issues which might be mass manufactured.<\/p>\n Photo voltaic panels are mass manufactured. Batteries, the issues that go in your telephone or in your electrical automobile, are 99% cheaper than they have been 20 years in the past. That\u2019s as a result of we manufacture them at a large scale. Each time you double the dimensions of producing, you get a ten% to twenty% discount in price, and there are such a lot of different issues like that.<\/p>\n On this case, the ability electronics, the transformer, are all special-purpose. Heron Energy goes to make the identical field for a knowledge heart, for an EV automobile charger, and for a photo voltaic farm. It\u2019s the identical field. No modifications. That\u2019s how we\u2019re going to get a price curve down for this stuff. That’s the most enjoyable pattern beneath this: the concept generalization and mass manufacturing of issues permits you, yr over yr, to scale back prices.<\/p>\n Whenever you\u2019re competing within the energy business, fossil gas prices have been principally stagnant. They go up and down a bit of bit, however in case you common them over 50 years they don’t seem to be on a cost-down curve. It doesn\u2019t get cheaper to get oil out of the bottom. My competitors is flat, and I\u2019m getting 10% to twenty% cheaper yearly. That\u2019s a fantastic enterprise to be in. That\u2019s the large pattern behind all of this. We noticed it first in photo voltaic and in batteries, however it allows an entire bunch of different issues in different industries, like energy electronics and extra.<\/p>\n Are we presently very depending on China for mass manufacturing?<\/p>\n Schroepfer: So much is coming from China, however I go to a manufacturing facility every week in the USA that’s getting spooled up with robotics, with actually sensible founders from Tesla and SpaceX. It seems that whenever you begin in 2026, you possibly can construct a way more environment friendly, a lot sooner manufacturing facility. You should use fashionable applied sciences.<\/p>\n Proper now, China has the commercial base, and we\u2019ve let it go. However I believe we’ve got a shot at rebuilding it in the USA, and I see good founder after good founder operating at this downside inside the USA on daily basis and each week.<\/p>\n It\u2019s one of many causes I began this agency, too. I believe we’ve got a shot to rebuild that industrial base in a next-generation set of know-how. Similar to areas world wide that didn\u2019t have landlines went straight to cellphones, we\u2019re going to go straight to totally automated robotic factories with 3D printing, laser milling and the most recent know-how set. It’s not going to be a cut-and-paste of what occurred in China, however a next-generation set of applied sciences that permit the U.S. to be self-sufficient in what we\u2019re doing.<\/p>\n We\u2019ve seen new methods. For example, uncommon earths have been one thing nobody ever talked about. Neodymium is that this uncommon earth materials that’s key to creating a magnet. Who cares about magnets? Properly, magnets are in each electrical motor in something. Something that has an electrical motor, you care about magnets. Virtually all of the neodymium is made in China, and it’s made on this very polluting, harmful course of. You do not need to go to considered one of these factories with fluorinated gases \u2014 it\u2019s terrible.<\/p>\n We\u2019ve obtained an organization making neodymium in Alameda, California. That’s not a simple place to allow polluting issues, which is okay for them as a result of their course of doesn\u2019t pollute in any respect. It\u2019s quite simple. It\u2019s two reactors. I walked across the facility. You don\u2019t want any protecting gear. As a result of it\u2019s so easy, they’re cost-competitive with Chinese language imports.<\/p>\n To their prospects who’re saying \u201cI\u2019m making an attempt to make magnets,\u201d they\u2019re saying \u201cnice, I’ll promote you neodymium. I’ve it. It\u2019s cost-competitive.\u201d<\/p>\n Everybody is happy, however the factor we\u2019re whispering within the background is, it\u2019s additionally not polluting. That is how we\u2019re going to win. It\u2019s not a cut-and-paste of that know-how over right here, however saying, \u201cHow will we strategy this in a means that\u2019s less complicated and cheaper, after which possible cleaner as properly?\u201d<\/p>\n We\u2019re doing the identical factor in copper. We\u2019ve obtained an entire bunch of bets in several sorts of supplies the place I believe we will do it higher within the U.S. We\u2019ve obtained an organization, Pink Metals, in South Carolina that\u2019s doing this for copper recycling. We\u2019re doing it in cement manufacturing. There’s a complete number of alternatives. I don\u2019t have sufficient time to satisfy all these entrepreneurs.<\/p>\n We talked rather a lot about a number of the corporations within the vitality sector. What are the opposite areas of investments that you just\u2019ve made that you just\u2019re enthusiastic about?<\/p>\n Schroepfer: I discussed this a bit, however price going a bit of deeper on is purposes of AI to the bodily world. I talked about one: Fractile, which is constructing a next-generation AI inference chip that\u2019s way more energy environment friendly.<\/p>\n One other instance is an organization referred to as Nyquis, which is utilizing AI to place a easy piece of {hardware} on an influence line, on each side of an influence line, to detect if there’s a fault in that energy line that is likely to be inflicting a hearth. The thought is that in case you detect that fault sooner, you possibly can stop the hearth earlier than it\u2019s an issue as an alternative of ready for it to occur after which having to reply. Utilizing AI plus {hardware} to determine this stuff out is one other instance of that.<\/p>\n We’ve got one other firm referred to as Vor Programs that\u2019s utilizing AI to assist with the nuts and bolts of how individuals make transactions to construct vitality initiatives. There may be quite a lot of due diligence work and different issues that must occur. You possibly can construct, very very similar to Harvey for authorized or OpenEvidence for medical doctors, these vertical AI corporations. It is a vertical AI firm for vitality builders. There’s a lot to occur there.<\/p>\n Then Rhoda is doing industrial automation with robots, utilizing next-generation fashions to coach robots to be simpler in manufacturing facility environments, again to my level of how we’re going to do that within the U.S. with superior robotics. I believe AI for the bodily world is a giant space.<\/p>\n I talked a bit about supplies: neodymium, copper. We’ve got an organization referred to as Dioxycle that\u2019s making clear chemical substances. These could be the large areas I’d spotlight.<\/p>\n I do know there are quite a lot of traders that you just associate with or work with which might be equally centered on this space.<\/p>\n Schroepfer: The factor that\u2019s been most fascinating is that there’s a set of oldsters who’ve been doing arduous tech or local weather for some time, and they’re nice companions of ours, from Congruent Ventures to Capricorn to Breakthrough Vitality to many others. However what\u2019s been fascinating to me is the generalist companies coming in. A quite common co-investor for us is Founders Fund, a16z, or Sequoia Capital. We\u2019re seeing them are available in giant quantities, as a result of they\u2019ve seen the financial alternative right here.<\/p>\n What did you study from spending 14 years at Meta?<\/p>\n Schroepfer: I realized a couple of issues. Once I joined in 2008, the corporate had fewer than 100 million customers, was not worthwhile, and had a few 150-person engineering group. We relied on exterior events to do all of the {hardware} work. We have been leasing information heart house.<\/p>\n Over the subsequent 14 years, we grew dramatically in customers and profitability and within the measurement of the group. However we additionally moved into the bodily world. As I confirmed you the server, we constructed our first information heart in 2011. I constructed 10 million-plus sq. toes of information facilities in 17 locations everywhere in the world. We then moved to shopper {hardware}, so we constructed the Ray-Ban sensible glasses, the Oculus Quest VR headset, and the Portal. Then we moved into AI analysis with the Fb AI Analysis Lab in 2013.<\/p>\n That shift into the bodily world introduced quite a lot of actually humbling classes. There have been quite a lot of instances the place stuff simply went incorrect. On the very first information heart, I keep in mind touring it underneath building, and we had wooden blocks on the loading dock as a result of that they had graded the loading dock incorrect, so the vans couldn\u2019t again up and unload correctly.<\/p>\n It\u2019s this new, superior, state-of-the-art information heart with a free-air cooling system, and we obtained incorrect the factor that each Walmart within the nation has 5 of. It\u2019s 1,000,000 small challenges.<\/p>\n That is the factor I deliver to the founders that I see: having realized learn how to construct stuff within the bodily world builds an appreciation for the dangers and scale, and for the way that you must emphasize pace and studying price.<\/p>\n Folks study the incorrect lesson. They assume {hardware} means spending quite a lot of time designing on paper. Improper lesson. You need to get on the market since you don\u2019t know which half goes to blow. You need to get on the market and study as quick as you possibly can and as cheaply as you possibly can, in order that whenever you\u2019re in mass manufacturing, you\u2019re not studying issues, you\u2019re simply repeating.<\/p>\n That lesson, from information facilities to shopper {hardware}, issues. Once we construct shopper {hardware}, you spend 18 months constructing this beautiful pair of glasses or this beautiful headset, however earlier than you promote it, you must do that drop-test factor, the place you actually say, what occurs when somebody takes it out of the field at dwelling and drops it on the bottom? If it breaks, they return it, and we eat the associated fee. So that you sit there and drop this factor with high-speed cameras again and again to verify it’s going to survive a drop from head peak. You don\u2019t consider this stuff whenever you\u2019re designing it. You need to be certain that somebody can drop it and it\u2019s wonderful, or spill some wine on it and it\u2019s wonderful.<\/p>\n These issues in the actual world, plus the challenges of constructing an govt group and scaling a corporation, are the enjoyable a part of my job: working with our founders and having their again when issues are robust, when they should recruit somebody, or once they\u2019re operating right into a problem in the actual world, as a result of I\u2019ve seen it. I\u2019ve seen all of it.<\/p>\n What\u2019s your response to the SpaceX IPO?<\/p>\n Schroepfer: I\u2019m actually fairly enthusiastic about it, as a result of we’ve got quite a lot of SpaceXers in our portfolio. I’ve quite a lot of mates who’re alumni or work at SpaceX. Having extra individuals on this planet with the monetary assets to work on audacious engineering initiatives goes to be actually good.<\/p>\n I believe it\u2019s additionally a lesson in constructing and {hardware}. What number of corporations can land rockets the best way SpaceX can? They\u2019ve been doing this for a decade, so that they have a really giant technical moat by way of what they\u2019re capable of deploy on this planet. Starlink is one other instance. Everyone seems to be racing to catch up. When you\u2019ve ever used Starlink on an airplane, you don\u2019t ever need to be on an airplane with out Starlink. It\u2019s arduous to explain different corporations which have such a singular product as SpaceX. I believe it\u2019s thrilling that the markets are rewarding that. I can\u2019t wait to see what SpaceX alumni do subsequent.<\/p>\n I think about there\u2019s going to be quite a lot of firm formation popping out of that IPO.<\/p>\n Schroepfer: It\u2019s going to be an thrilling 5 years. I met you after I began my first firm in 2000 and offered it off. We checked out beginning one other firm, after which I labored at Mozilla and Fb, so I\u2019ve been by means of a pair cycles of this. I believe it’s the most enjoyable time to begin an organization, by way of the capital out there, the AI instruments out there to you, and the bodily instruments to construct issues shortly within the bodily world. It\u2019s the wager I made: I believe that is probably the most fascinating time to be constructing new corporations. That\u2019s the smaller model of why I did this. I believe that is the time. That is the factor to be doing.<\/p>\n Illustration: Dom Guzman<\/p>\n![\"\"](\"https:\/\/news.crunchbase.com\/wp-content\/uploads\/Mike-Schroepfer-Gigascale-Capital-Speaking-Candid-BW-2026-1-300x255.jpeg\")
![\"\"](\"https:\/\/news.crunchbase.com\/wp-content\/uploads\/Gigascale-Capital-Founders-Team-Photo-2026-scaled-e1782247124286-300x177.jpeg\")
\u2018Not software program with increased capex\u2019<\/h2>\n
![\"\"](\"https:\/\/news.crunchbase.com\/wp-content\/uploads\/Panthalassa-Ocean-Platform-In-Operation-2026-scaled-e1782247451404-300x235.jpeg\")
![\"\"](\"https:\/\/news.crunchbase.com\/wp-content\/uploads\/Rhoda-AI-Robot-Folding-Clothes-2026-300x164.jpg\")
Associated Crunchbase queries:<\/h2>\n
![](\"https:\/\/news.crunchbase.com\/wp-content\/uploads\/energy-tech.jpg\")
<\/p>\n