About
I like to build!
Articles by Lior
Activity
34K followers
-
Lior Susan posted thisStop Talking About Gross Margins I still see investors obsessing over gross margins. This needs to stop. Gross margin is a vanity metric. It tells you almost nothing about whether a business is actually creating value. What matters is free cash flow - how much real money the business generates after everything is paid for. A company with 80% gross margins and massive R&D burn, bloated sales teams, and endless customer acquisition costs can destroy value for a decade. I've watched it happen. A company with 40% gross margins that converts efficiently to free cash flow, owns its supply chain, and has customers who can't leave? That's a machine. This is especially true in physical industries - the sectors we build in at Eclipse. Manufacturing. Defense. Energy. Hardware. These businesses will never have software margins. They shouldn't. And that's fine. Here's what I care about: Free cash flow conversion. How much of every dollar of revenue turns into actual cash? A 40% GM business converting 20% to FCF is better than an 80% GM business burning cash to grow. Customer lock-in. In physical industries, switching costs are real. When you're embedded in a factory floor, a defense program, or an energy grid - you don't get replaced by a cheaper alternative next quarter. That durability matters more than margin percentage. Pricing power. Can you raise prices without losing customers? Defense contracts, industrial equipment, critical infrastructure - these have real pricing power that SaaS companies dream about. The best businesses I've seen in the physical world have gross margins between 35-55%. They also have 20-year customer relationships, massive barriers to entry, and generate serious free cash flow. Stop screening out the future because it doesn't look like a SaaS P&L. The investors and operators who understand this will build the next generation of great companies. The ones still worshipping gross margins will wonder what happened with their multiples.
-
Lior Susan posted thisEveryone talks about building legendary companies. Very few talk about surviving long enough to become one. The truth is, most companies don’t fail because the vision was wrong. They fail because they ran out of time, cash, or discipline on the way to that vision. Don’t die on your way to glory!!! We romanticize bold bets, massive scale, and “go big or go home.” But the real game is more nuanced: - Stay alive while you figure it out - Stay funded while you scale it - Stay disciplined while others chase headlines Because if you die at mile 20, it doesn’t matter how right you were. The best founders I know understand this deeply: Survival is not weakness - it’s strategy. They manage burn like it’s oxygen. They sequence risk instead of stacking it. They know when to push… and when to endure. Especially now - in a world of capital cycles, geopolitics, and real-world execution -building matters more than ever. But none of it matters if you don’t make it to the other side. Build for glory, Operate for survival.
-
Lior Susan shared thisAmerica’s ‘Featherweight’ Missile Is Unlike Anything China Has. It Could Rewrite the Hypersonic Arms Race. Ursa Major!Lior Susan shared thisPopular Mechanics Magazine just published a deep look at HAVOC, the hypersonic missile system built by Ursa Major. The U.S. is taking a fundamentally different approach. HAVOC is lean, precise, and launchable from fighter jets, bombers, ground systems, and space. No nuclear warhead. Pinpoint accuracy. The engine changes everything. Ursa Major's Draper runs on storable propellants — hydrogen peroxide and kerosene. No cryogenic fueling, no boil off, no launch windows. A HAVOC missile can sit loaded on a truck, a wing pylon, or a silo for months and fire on command. That transforms the logistics of hypersonic warfare from a planned event into a persistent, ready capability. You go from "schedule the launch" to "pull the trigger." And because Draper is throttleable from 10% to 100% thrust, the operator has real time control over the flight profile. That means active trajectory shaping, terminal maneuvering, and the ability to adjust speed and vector mid-flight to defeat air defenses. Solid rockets give you one burn profile. Draper gives you a flight envelope. That is the difference between a ballistic projectile and a weapon that thinks on the way to its target. They flight tested it this month, concept to supersonic flight in 8 months, and the first attempt was flawless. And critically, Draper is designed to be manufacturable at scale. Not a lab prototype. A production engine. The U.S. doesn't need bigger missiles. It needs smarter, cheaper, more producible ones. HAVOC is that and we are one TRL 7 flight from an F15 from changing munitions forever. https://lnkd.in/gF2GubGZAmerica’s ‘Featherweight’ Missile Is Unlike Anything China Has. It Could Rewrite the Hypersonic Arms Race.America’s ‘Featherweight’ Missile Is Unlike Anything China Has. It Could Rewrite the Hypersonic Arms Race.
-
Lior Susan posted thisThe Biggest Industrial Opportunity Nobody's Talking About The Middle East is sitting on the fastest GDP growth story of the next decade. But it's locked behind one regime. Iran has 90 million people, a young and educated population, massive natural resources, and a geographic position connecting Asia, Europe, and Africa. Its neighbors - Saudi Arabia, the UAE, Qatar, Bahrain - are already spending hundreds of billions on AI infrastructure, advanced manufacturing, and energy transition. But none of it reaches its full potential while Iran exports instability instead of innovation. I grew up in the Middle East. I served in its military. I've watched brilliant people across the region build incredible things despite the constraints. And I've spent the last decade building at the intersection of technology and the physical world through Eclipse Here's what I see: If Iran's regime changes, the entire region unlocks. Not in a decade. In years. Here's why: 1. The capital is already there. Gulf sovereign wealth funds control trillions. Abu Dhabi, Saudi Arabia, and Qatar are already deploying into AI, data centers, advanced manufacturing, and defense. Remove the security overhang and that capital accelerates - into Iran, into cross-border infrastructure, into regional supply chains that today don't exist. 2. The talent is there. Iran produces more engineers per capita than almost any country in the region. Add that to the UAE's ambitions, Saudi's Vision 2030, and Israel's deep tech ecosystem - you have a talent corridor from Tel Aviv to Tehran to Dubai that could rival Silicon Valley in a generation. 3. The energy is there. The region has sun, gas, oil, and increasingly nuclear. Industry needs power. Manufacturing needs power. The Middle East has it in abundance - and the will to deploy it. 4. The geography is there. Sitting between the world's largest consumer markets (Asia) and its wealthiest (Europe). A stable Middle East becomes the manufacturing and logistics bridge for a rewired global supply chain. What's missing isn't capital, talent, energy, or geography. It's stability. And stability starts with Iran. A post-regime Iran doesn't just change Iran. It changes everything around it. Israel normalizes with more neighbors. Saudi Arabia trades instead of arms. The Gulf states stop spending on defense and start spending on growth. New trade corridors open. New industries emerge. The technologies we build at Eclipse - autonomous factories, smart energy grids, advanced manufacturing - are exactly what the region needs to industrialize at speed. The Middle East doesn't need more sovereign wealth funds. It needs regime change in Tehran and an industrial revolution.
-
Lior Susan shared thisIt was fun to work on this one! Today, Amazon Web Services (AWS) and Cerebras announced a collaboration to bring ultra-fast inference co-powered by Cerebras’ CS-3 systems and AWS’ Trainium-powered servers to Amazon Bedrock. The collaboration represents an “aha” moment – as AI continues to proliferate and workloads become even more diverse and demanding, heterogeneous AI is becoming mission critical. That’s why I knew this partnership made sense! The partnership has game-changing potential, especially for the organizations operating within the AWS environment; organizations we all rely on every day. Congratulations to both teams. This is the future of compute.Lior Susan shared thisToday we announced a game-changing collaboration between Amazon Web Services (AWS) and Cerebras to deliver the fastest AI inference available for generative AI applications. Inference is where AI actually delivers value to customers. As models become more capable and reasoning workloads explode, inference speed has become one of the biggest bottlenecks holding back AI innovation. Together with Cerebras, we’re reimagining inference architecture on Amazon Bedrock. We’ve built a disaggregated system that separates prompt processing from output generation. AWS Trainium3 powers the prefill stage, while the Cerebras CS-3 system accelerates decoding. High-speed AWS networking seamlessly connects the two, allowing each processor to do what it does best. The result is dramatically faster inference performance for generative AI and large language model workloads. This breakthrough capability launches on Amazon Bedrock in the next few months — and AWS is the exclusive cloud provider for Cerebras's disaggregated inference solution. Learn more: https://lnkd.in/dRgS5XWV
-
Lior Susan shared thisWe incubated this company at Eclipse and Rivian a few months back with RJ Scaringe, based on our manufacturing expertise and the opportunity to transform these industries through robotics and AI. https://lnkd.in/gTaqUHuuExclusive | Rivian CEO’s AI-Powered Robotics Startup Raises $500 MillionExclusive | Rivian CEO’s AI-Powered Robotics Startup Raises $500 Million
-
Lior Susan shared thisThere’s a reason the Ayatollah’s regime directs so much of its hostility toward the UAE. The UAE represents everything that fundamentalist regimes fear most: a successful, modern country that is open to the world, embraces innovation, builds bridges across cultures, and maintains strong relationships with the United States, Israel, and many partners across the globe. From my own experience working in the region, the leadership and community in the UAE are extraordinary. His Highness Sheikh Mohammed bin Zayed, His Highness Sheikh Tahnoon bin Zayed, and institutions like G42 are building a vision of the Middle East centered on progress, technology, cooperation, and opportunity. The UAE shows that the region’s future does not have to be defined by extremism or isolation. It can be defined by growth, openness, and partnership with the world. That is precisely why it draws so much fire from those who fear that model. Despite the attacks and rhetoric, the UAE stands as one of the most hopeful examples of what the Middle East can become — alongside partners who share the same vision for stability, innovation, and peace. Picture I took from Stargate UAE 2 weeks ago!
-
Lior Susan posted thisOperators With Capital There's a reason most firms struggle in the physical world. It's not the deal flow. It's not the capital. It's that they don't know how anything gets built. When a founder walks into a pitch meeting and talks about yield rates, qualification cycles, or ITAR compliance — most investors nod politely and Google it later. They can model a TAM. They can't tell you why a production line is down at 3 AM or what it actually takes to get a defense contract from prototype to production. That's the gap. And in the physical world, that gap kills companies. I. Software is forgiving. You ship fast, break things, patch it Friday. The feedback loop is measured in hours. Hardware is unforgiving. A manufacturing defect doesn't get a hotfix. A missed qualification window costs you a year. A supplier going sideways doesn't mean a bad quarter — it means your company might not exist next quarter. The founders building in this world don't need another investor who "adds value" through warm intros and board decks. They need someone who's actually stood on a factory floor. Someone who's scaled a production line from ten units to ten thousand. Someone who knows what it feels like when the machine breaks and the customer is waiting. They need operators. II. At Eclipse, every partner has built something with their hands before they ever wrote a check. This isn't a recruiting pitch. It's the point. When our founders call at 2 AM because their supplier just fell through, we don't connect them to an advisor. We get on the phone and help solve it — because we've been there. When a portfolio company needs to rethink their production architecture, we don't hire a consultant. We sit in the room and draw it on the whiteboard — because we've done it before. Capital is table stakes. Anyone can write a check. The question is: what happens after? III. The world is waking up to something we've believed for a decade: the physical world is where the biggest opportunities are. AI needs chips. Chips need fabs. Fabs need power. Power needs infrastructure. Defense needs manufacturing. Manufacturing needs automation. Every layer of that stack is a physical problem. And physical problems require people who understand physics, supply chains, materials, production, and scale. Not spreadsheet operators. Real operators. IV. If you're a founder building something you can touch — something hard, something physical, something the world actually needs — find investors who've built things too. The best firms in the physical world won't be built by financiers who learned manufacturing from a podcast. They'll be built by operators who learned finance on the job. That's the bet we made. And the world is proving it right.
-
Lior Susan posted thisThe Industrial Moment For a decade, the most celebrated companies in tech made things you couldn't touch. Software. Ads. Feeds. The returns were great. The products were weightless. Meanwhile, the stuff that actually holds civilization together — chips, engines, medicines, energy, ships — got offshored, hollowed out, forgotten. We called it efficiency. We called it globalization. It wasn't progress. It was a bet that the world would stay stable and someone else would always make the hard stuff for us. That bet is over. I. Look around. Not at pitch decks. At the world. The Strait of Hormuz is contested. Semiconductor supply chains run through one island 100 miles off China's coast. Europe is rearming for the first time in a generation. The US can't produce munitions fast enough. We can build the most sophisticated AI models on earth — and can't make enough chips to run them. This isn't a cycle. It's a reckoning. The companies that will define what comes next aren't building another dashboard. They're building propulsion systems, autonomous vehicles, cell therapies, rocket engines, factory robots. The hard stuff. The stuff you can touch. II. We started Eclipse in 2015 on a simple belief: the biggest opportunities in technology are in the physical world, and the best people to back them are operators who've actually built things. For years, that was contrarian. The smart money chased software margins. We were in factories and labs, backing founders who bent metal and filed patents measured in nanometers. People called it hard tech. We just called it the future. Today that future is showing up — not on a roadmap, but under pressure. National security, supply chain fragility, energy demand, AI's hunger for physical infrastructure — all converging into the largest industrial investment cycle since World War II. III. A few things I believe: The AI revolution is a physical revolution. Every frontier model needs chips, power, cooling, data centers. The companies that build and maintain this infrastructure will capture as much value as the models. Maybe more. Manufacturing is a strategic weapon. The country that can make things fast and at scale has leverage over the one that can only design them. China figured this out 40 years ago. Defense is no longer a niche. The founders leaving SpaceX, Tesla, Anduril are building companies that blur commercial and defense. The Pentagon's budget is a market signal. Capital alone isn't enough. Founders building physical products don't need another board member who forwards intros. They need partners who've run production lines and stared down a yield problem at 2 AM. IV. To the founders building the hard stuff: This is your moment. Not because it's easy — it's brutally hard. But the world finally needs what you're building, and it needs it now. Build things that matter. Build things you can touch. The Industrial Moment is here. Don't waste it.
Lior Susan commented on a post
18h
Nothing beat a good Eclipse! :)
Lior Susan commented on a post
1w
Brad Svrluga Unfortunately, with this mayor, this is only the beginning. The American Dream that brought me to this country doesn’t seem to be part of this person’s vision or agenda. Keep up the fight!
-
Lior Susan liked thisLior Susan liked thisThere is a fundamental misconception that to lead, people must work for you. This is wrong. And profoundly backwards. We don't promote people into leadership. We generally promote people who are already leading. In your clubs, your sports teams, your study groups, at work, at school, at church, temple or mosque… you learn to lead by leading. The best leaders I know do exactly that. They make their peers' jobs easier, not harder. They do the things no one wants to do, but that need to be done. They don't complain. They don't make excuses. And people choose to follow them. How do you learn to lead? The answer is deceptively simple. Leading is a muscle like any other muscle. You learn by repetition. Nobody needs to anoint you. You don't need to be the captain or the boss. Start leading now and the title will catch up.
-
Lior Susan liked thisLior Susan liked this#1 in the Energy sector for Most Innovative Companies?! That’s us at Redwood Materials! Couldn’t be more proud! https://lnkd.in/eZEUAE9P
-
Lior Susan liked thisLior Susan liked thisFor the first time ever... we're revealing how Mashgin actually works. Before you get any ideas, this is patent protected technology and you couldn't copy it if you tried.
-
Lior Susan liked thisLior Susan liked this#XoraPortfolio company Peak Energy has been named to TIME’s Top GreenTech Companies list! In just a few years, Peak has grown into a company focused on a clear challenge: lowering the cost of energy storage. Its work on passively cooled sodium-ion systems and large-scale grid deployments points to a broader shift toward storage solutions designed for cost, safety, and scalability. Congrats Landon Mossburg, Cameron Dales, Liam O'Connor, Geoff Brown, Paul Durkee, Brandon Kelly and the full Peak team for this achievement! Full list: https://lnkd.in/eQ-jKYE7 #PeakEnergy #SodiumIon #BESS #XoraInnovation
-
Lior Susan liked thisLior Susan liked thisChristopher Spagnoletti, CEO of Ursa Major, delivered the industry keynote at the Defense Manufacturing Conference (DMC) this week in Orlando. His message on deterrence was clear--innovation alone doesn’t win. Deterrence is built through scale on factory floors, on test stands, and in the ability to produce reliable systems at the speed and volume the warfighter requires. At Ursa Major, we’re focused on delivering both advanced capability and the industrial base to scale it. #FlyFaster #DMC25
-
Lior Susan liked thisLior Susan liked thisHiring full-stack SWE's in beautiful 🌲 🌞 🏔️ North Vancouver to come work on electrifying the continent. Come join one of the most talented SW teams working in clean tech and knock down emissions one home at a time. https://lnkd.in/gNV9vSFq
-
Lior Susan liked thisHmmmmmmm. I smell new warehouse smell. Must be something new coming. But where? How long does it take to count to 16M ? Announcement coming sooner than it takes to do that. Maybe a few more.
-
Lior Susan liked thisI had a fun conversation with Tony Hoang of the AI Podcast about the rapid growth of inference workloads, the tech we’re building at Cerebras to power this growth, why NVIDIA has no moat with CUDA, and lastly why we’re still at the early days for AI adoption. Check it out!Lior Susan liked thisJoin Hagay Lupesko, SVP of AI Inference at Cerebras Systems, for a deep dive into the rapidly evolving world of AI inference. Hagay breaks down why inference has overtaken training as the dominant AI workload, how Cerebras' wafer-scale chip architecture delivers 10-20x faster performance than NVIDIA GPUs, and why CUDA is no longer the moat many think it is. He also covers how DeepSeek wiping $600 billion off NVIDIA's market cap in a single day was both a foundational and deeply misunderstood moment for the industry, the growing energy crisis in AI infrastructure, and what it will take to support the explosive rise of AI agents in the enterprise. Listen here: https://lnkd.in/e7vRNWCFInterview #84 Hagay Lupesko, SVP of AI Inference at Cerebras SystemsInterview #84 Hagay Lupesko, SVP of AI Inference at Cerebras Systems
Experience & Education
-
Eclipse
View Lior’s full experience
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
Languages
-
English
-
-
Hebrew
-
Other similar profiles
Explore more posts
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content