Convincing on the physics and the cost curve. The question that decides whether SSTs matter this decade is timeline to scale: SiC supply, utility certification cycles, integration with existing protection schemes, and cybersecurity sign-off are all gating items measured in years, not quarters. With grain-oriented electrical steel in 30% deficit and conventional transformer lead times now 3-4 years, the real comparison isn't SST vs. legacy transformer on a spec sheet — it's whether SST-at-scale arrives before or after the next 200 GW of data center load tries to interconnect.
I also listened to the podcast with Shayle. It's a powerful pitch.
This is the same problem India will hit at scale. You can build generation fast — solar panels arrive in containers. What you cannot build fast is the transmission corridor, the substation, the interconnection queue, and the high-voltage transformer. Generation is the visible layer. Delivery infrastructure has a 5-to-10-year lag. That gap is where supercycles stall.
The stat that jumped out to me: inverter failures accounting for 36% of lost energy in utility-scale solar. If SSTs can meaningfully dent that number through built-in redundancy and real-time monitoring, the ROI case basically writes itself for developers who are already bleeding margin on O&M.
I can't understand why founders given the opportunity to guest post on the a16z Substack choose to post AI slop. Do you think we can't tell??
Seems a waste of a great opportunity to attract top-tier talent and raise your company's profile.
I really enjoyed your newsletter. Thank you.
I had a few questions about what you wrote, and I was wondering if it would be okay to ask you personally.
Would you be willing to send me an email? I can reply with my questions.
davidkimxlab@gmail.com
Or if you prefer, please let me know your email address and I will reach out first.
I would really appreciate your reply.
https://sparks.egnyte.com/dl/XFyfRhjmqJ6x
e2 Companies looking for power in all the wrong places solving the AI Data Center problem without breaking the grid.
Convincing on the physics and the cost curve. The question that decides whether SSTs matter this decade is timeline to scale: SiC supply, utility certification cycles, integration with existing protection schemes, and cybersecurity sign-off are all gating items measured in years, not quarters. With grain-oriented electrical steel in 30% deficit and conventional transformer lead times now 3-4 years, the real comparison isn't SST vs. legacy transformer on a spec sheet — it's whether SST-at-scale arrives before or after the next 200 GW of data center load tries to interconnect.
I also listened to the podcast with Shayle. It's a powerful pitch.
This is the same problem India will hit at scale. You can build generation fast — solar panels arrive in containers. What you cannot build fast is the transmission corridor, the substation, the interconnection queue, and the high-voltage transformer. Generation is the visible layer. Delivery infrastructure has a 5-to-10-year lag. That gap is where supercycles stall.
The stat that jumped out to me: inverter failures accounting for 36% of lost energy in utility-scale solar. If SSTs can meaningfully dent that number through built-in redundancy and real-time monitoring, the ROI case basically writes itself for developers who are already bleeding margin on O&M.