The Wood Mackenzie report landed August 13, the eve of Independence Week, with a yummy number: 34 gigawatts. That is how much solar capacity we installed in the first six months of 2026 alone! 34 GW sounds abstract, so lets do the math: At a 20% capacity factor - generous but not far off what Indian solar projects are getting in Rajasthan and Gujarat - 34 GWdc translates to roughly 60 billion units of electricity per year. Rajasthan’s entire residential electricity demand is somewhere in that neighbourhood. We built, in six months, a solar fleet capable of lighting the households of an entire state.
And that was not even the most important thing that happened this week.
The Sun Is Not Stopping
India is on track to add more than 50 GW of solar in calendar year 2026, according to both Wood Mackenzie and BNEF. If that holds, we will overtake the United States to become the world’s second-largest solar market. Until very recently, the United States was the reference point - the thing you compared your numbers against to feel significant. We are no longer comparing. We are passing.
The H1 2026 figure of 34 GWdc represents a 38% jump over H1 2025, when we added roughly 24.6 GWdc. The second half is expected to slow - domestic cell manufacturing capacity hasn’t kept pace with installation demand, and module prices have ticked up - but even a flatter H2 produces a full year that would have seemed implausible two years ago.
The grid consequences are real. Somewhere in our transmission network, substations designed for a certain load profile now have more input than the engineers originally planned for. That is a procurement and grid management problem we are increasingly being forced to solve, and solving it is pushing a generation of power engineers into genuinely creative territory. Our grid challenges are now first-world problems.
Alongside solar, the private nuclear race started in earnest this week. Indian Infrastructure Magazine confirmed that six of the largest industrial conglomerates in the country - Hindalco, Jindal, Tata Power, Reliance, JSW Energy, and Adani Power - have each submitted responses to an expression of interest for private nuclear participation.
Tata Power is planning a 440 MW Bharat Small Modular Reactor in collaboration with NPCIL. Adani Power has targeted 10 GW of nuclear capacity by 2035 and incorporated dedicated subsidiaries to hold nuclear licenses.JSW Energy wants its first nuclear plant operational by 2030.
The scale ambition is 100 GW of nuclear capacity by 2047. We currently have roughly 8.8 GW. That is 91 GW more than today - twelve times our current nuclear base, in two decades. I do not know if it is achievable; the lead times on large reactors make it a very long shot unless the SMR program delivers on schedule, and Indian infrastructure programs have a history of ambitious targets meeting complicated ground conditions.
But I do know that six of the largest private capital pools in the country have decided to find out, and Rs 200 billion is already allocated in Budget 2025-26 for SMR design, development, and deployment. Five indigenously designed SMRs are targeted for operation by 2033. I wrote earlier this year about the physics and the 70-year institutional bet underneath all of this and the logic of that piece has aged well. The private sector entrants are the new variable.
The Chips Are Finally Coming Home
IANS reported that seven chips have now been fabricated domestically in India, including at an advanced 12 nanometre node. This is evidence that the infrastructure works: the design tools, the process flows, the cleanroom environment, the packaging chain. The question of whether India can do this has been answered, at small scale. The question of whether we can do it at a scale that matters economically continues to remain teasingly open.
But the pipeline feeding that question is substantial. Twelve semiconductor projects have been approved across six states, with a total investment pipeline exceeding Rs 1.64 lakh crore. 211 chips have been taped out by 75 Indian institutions as of April 2026, covering defence electronics, communications, and automotive applications.
The compute stack above the silicon is filling in quickly: the IndiaAI Mission’s shared platform has reached 45,000+ GPUs, and 237 projects have consumed 93.18 lakh GPU-hours at subsidised rates through August.
The investment figures are also heartening: Rs 26.75 lakh crore in investment proposals filed with government authorities between April 1 and August 5 of FY 2026-27, per Bank of Baroda Economics Research.
Of that, Rs 14.98 lakh crore - roughly 56% - is in ITeS, and 99% of that ITeS pile is pointed directly at data centres and AI projects.
India added 258 MW of data-centre capacity in H1 2026 alone, a 59% jump over the same period in 2025, bringing our total operational data-centre capacity to approximately 1.8 GW. Nuclear power is the second-largest category in that investment pipeline: Rs 6.5 lakh crore from four companies. The energy transition and the compute transition are so tightly intertwined now that treating them as separate stories is quickly becoming impossible.
Space Is Turning Commercial
The Rajya Sabha heard on August 13 that India’s space sector now counts 440 registered startups, up from 323 in earlier counts. Eighteen have received authorisations from IN-SPACe to conduct space activities - Agnikul Cosmos, Skyroot Aerospace, GalaxEye Space Solutions, Dhruva Space, Bellatrix Aerospace, Digantara, Pixxel, Manastu Space, and Space Kidz India among them.
The government told Parliament that private companies are expected to conduct two commercial launches in FY 2026-27, rising to more than six in FY 2027-28.
Two launches sounds modest. Six starts to look like a market. I wrote about the structural transformation that made this possible here and here and the trajectory since those pieces has been faster than I expected.
The structural change is visible in the consolidation beginning inside the ecosystem. On August 10, GalaxEye - a multi-sensor Earth Observation startup - acquired StarOps, a Bengaluru spacecraft engineering firm founded in 2022. StarOps brings spacecraft systems engineering, propulsion, avionics, flight computing, guidance and navigation, and mission operations under GalaxEye’s roof, enabling the company to close the loop internally from sensor design to mission operations. That is a significant move: the global EO market is shifting toward vertical integration, and acquiring StarOps shortens GalaxEye’s supply chain considerably while adding capabilities that were previously contracted out.
Also worth noting, because it never quite makes headlines at the scale it deserves: Sarla Aviation, a Bengaluru eVTOL startup building electric air taxis, received a Design Organisation Approval from DGCA on August 10. This is the regulatory clearance that allows a manufacturer to move from prototype development to production planning. We are possibly years away from landing at Bengaluru’s beautifully bamboo-zled Terminal 2 and taking an eVTOL for a 20-minute flight to Indiranagar. Wouldn’t that be the day!