Tim Reichard @reichardtim - Twitter Profile

about 1 month ago

The more I dive deep into the solid-state battery space and the opportunity, the stronger the $QS @QuantumScapeCo thesis gets. I keep pulling threads and every single one leads back to the same place. Let me show you what I found. 🧵 First - the macro setup. Data centers now eat 5% of all US electricity. McKinsey says that doubles by 2030. Goldman says data center power demand grows 175% this decade - that's like adding an entire top-10 country to the global grid. And we're already failing. Bloomberg reported 30-50% of planned 2026 US data centers are delayed or canceled. Not because of money. Not chips. Because we literally can't plug them in. @KobeissiLetter nailed it: "Energy will soon be the AI bottleneck." It already is! So what's the fix? Follow what @nvidia is actually doing. @nvidia is building 800V DC power architecture for next-gen AI factories. 1 MW per rack. Shipping with Kyber in 2027. They've pulled in Vertiv, Schneider, Eaton, TI, Infineon, STMicro - the entire power ecosystem. Why 800V DC? Traditional AC physically cannot push enough watts into a rack at AI-scale density. Enteligent's white paper showed DC-native racks generate 10-15x more revenue per square foot. TI unveiled a full 800VDC reference architecture at GTC 2026. This is happening now. Here's the part that gets me very excited! Vertiv - NVIDIA's primary power infrastructure partner - said their 800 VDC platform is being designed with INTEGRATED ENERGY STORAGE. Not backup. Not UPS. Integrated into the architecture. Battery storage. Baked into the rack. By design. So what battery do you put inches away from GPUs worth half a million dollars? → Energy-dense (every square inch is revenue) → Zero thermal runaway (lithium-ion fire next to your H100s = game over) → High cycle life at rack-level charge/discharge rates → Safe at 800V operating voltage Lithium-ion? Thermal runaway. Can't risk it next to the compute. Lead-acid? Not close on density. Flow batteries? Way too big. Solid-state checks every box. Now here's where it gets interesting - the technical moat. $QS isn't just "a solid-state battery company." Look at what they've actually built: → 300+ patents covering the ceramic separator and anode-free cell design. This isn't one patent. It's a fortress. → Anodeless architecture - zero excess lithium, no host materials. This is what enables the energy density advantage. → Their ceramic separator is the only one that's demonstrated dendrite resistance at automotive-grade current densities for 800+ cycles at room temperature. Nobody else has shown this. Not Toyota. Not Solid Power. Not CATL. → 1,000 cycles with 95% capacity retention. Published data, not a press release. → Cobra manufacturing process - 25x faster heat treatment, fraction of the floor space vs their prior Raptor process. This is how you go from lab to factory. → Murata AND Corning - two of the world's best ceramics companies are both investing in QS-proprietary tooling to produce the separator. When the two best in the world independently validate your manufacturability, that tells you something. Competitors? Solid Power is on sulfide electrolytes - different chemistry, different problems. Toyota says 2027-2028. CATL says 2027. BMW is still on prototypes. $QS has cells on Eagle Line today. The moat isn't just the tech. It's 300 patents + the only proven dendrite-resistant ceramic separator + a manufacturing process that actually scales + two world-class production partners already building the tooling. Good luck replicating that in 2 years. The recent hiring in 2026 confirms the thesis entirely! → VP/GM from @MediaTek's data center AI infrastructure group → @nvidia / @Cadence ex-CFO Geoff Ribar on the board → Ex-CTO of L3Harris and ex-President of Aerojet Rocketdyne on the board → Commercial Counsel with STEM background and licensing focus → Licensing product strategy role to define technology transfer value prop → @ironmantimholme (CTO) is publicly framing energy delivery as the next AI bottleneck They're not talking about pivoting. They already hired for it. They already have board governance for it. They're building the cells for it. Connect the dots: AI energy bottleneck is here → NVIDIA building 800V DC → architecture requires in-rack storage → solid-state is the only chemistry that fits → $QS has the deepest technical moat, 300+ patents, the only proven separator, and is already staffing for data center and defense commercialization Market still think this is an EV battery company. That's the edge. $QS

12

113

12

38

7K