How Tesla Plans to Scale Solar Production to 100GW by 2028

Last Updated: May 2026 • Based on Tesla Master Plan Part 3, Tesla Energy deployment reports, and independent analyst tracking (Wood Mackenzie, BloombergNEF)

In April 2023, Tesla published Master Plan Part 3 — a 41-page engineering and economics document that tries to answer a single question: what would it actually take to run the entire global economy on sustainable energy? The answer the document lands on is roughly 30 terawatts of renewable generation (mostly solar) and 240 PWh of storage. Buried inside that vision is the manufacturing number that matters to everyone in this trade: Tesla wants to be producing solar at a rate of 100 gigawatts per year by 2028.

Stop and think about that number the way we do at the supply-house counter. The whole planet installed somewhere around 450–500 GW of solar in 2023. Tesla is saying one company — a company that deployed about 4.1 GW of solar in 2023 — intends to manufacture a fifth of today's global annual volume on its own within five years. We've watched module makers ramp from 5 GW to 50 GW before; it took Jinko the better part of a decade. Tesla wants to do a 24× ramp in five years.

This guide breaks down how Tesla says it will get there: the Gigafactory manufacturing playbook, the Solar Roof versus panel two-track product strategy, why Powerwall and Megapack integration is the real moat, the supply-chain wall the plan has to climb, and — most important for our customers — what a Tesla at 100 GW means for installers, EPCs, developers, and equipment buyers between now and 2030.

Master Plan Part 3 is not a marketing deck. It's a capacity model. Tesla's engineers worked backward from a fully electrified global economy — all transport electric, all heat electrified, all industry electrified — and calculated the generation and storage hardware that economy requires. Their figure of roughly 30 TW of renewable capacity is dominated by solar because solar is the cheapest, most scalable, most universally available generation source on the planet.

The 100 GW/year target is Tesla's answer to the follow-up question: what share of that hardware does Tesla intend to build itself? At 100 GW per year sustained for a decade, Tesla alone would add a terawatt of solar manufacturing output — three percent of the 30 TW global requirement, from a single company. That's the ambition. Whether the factory construction, the polysilicon, the silver, and the labor all line up on that schedule is a different question, and we'll get to it.

I've sat in distributor meetings where manufacturers pitch five-year capacity roadmaps, and I've learned to discount them by a third and slide them two years right. Apply that rule of thumb to Tesla and you still get a company producing 60+ GW of solar a year by 2030 — which would make it one of the three largest module producers on earth. That is worth planning around whether or not the 2028 date holds.

Tesla Energy's solar deployment in 2023 was approximately 4.1 GW across residential Solar Roof and conventional panel installations. Growth accelerated through 2024 and 2025 as the Solar Roof line at Gigafactory Texas ramped, but even generous estimates put Tesla's current solar run rate in the 8–12 GW/year range — an order of magnitude below the target. On the storage side the story is much stronger: Tesla has deployed more than 25 GWh of Megapack cumulative capacity and the Lathrop, California Megafactory is producing at roughly a 40 GWh/year clip.

The asymmetry matters. Tesla's energy business is, today, primarily a storage company with a solar side hustle. The 100 GW plan flips that — and the table below shows how steep the required ramp really is.

For calibration, Tesla's vehicle production grew from about 500,000 cars in 2020 to 1.8 million in 2023 — a 3.6× ramp in three years, and that nearly broke the company twice. The solar plan demands a ramp more than six times steeper. Possible? Tesla has a habit of hitting the number late. But nobody in the supply chain should bank a project pipeline on the 2028 date.

The core of the plan is the same playbook that took Tesla from a boutique automaker to the most valuable car company on earth: build factories so large, so automated, and so vertically integrated that per-unit cost falls faster than competitors can chase with design tweaks alone. Applied to solar, that means five operational pillars:

• Vertical integration. Wafers, cells, glass, frames, and modules under one roof. Every supplier margin eliminated is two to four cents a watt off the stack — and in a market where Tier 1 modules trade at $0.20–0.35/watt, that's the whole ballgame.
• Robotics and automation. Tesla's vehicle lines taught it how to throw automation at assembly problems. Module assembly lines that used to need dozens of hands per shift can run with a fraction of the labor, which is how you make panels in New York or Texas at Chinese labor-cost parity.
• Co-location with storage. Building solar and battery lines in the same campus lets Tesla bundle hardware at the loading dock instead of at the job site — cutting freight, packaging, and channel cost on every solar-plus-storage system sold.
• Continuous process engineering. The iterative factory discipline behind the 4680 cell program applied to cell efficiency and line yield. Every half-point of efficiency gained is half a point of cost per watt removed without touching the bill of materials.
• Geographic diversification. Factories near demand centers cut freight cost, sidestep tariff exposure, and — critically in the US — qualify for IRA Section 45X manufacturing credits and domestic-content bonus stacking on the projects that buy the output.

Gigafactory New York in Buffalo, built under the old SolarCity/Panasonic partnership arrangement, runs at roughly 5 GW/year of Solar Roof and panel capacity. That is the proof of concept, not the plan. The plan requires four to five additional facilities at 20 GW or more each — the largest single-site solar factories ever attempted outside China.

Tesla is not trying to win one solar market. It's running two products at two different markets with two different moats, and understanding the split is the key to understanding where 100 GW of volume could actually come from.

The Solar Roof is the premium play. Roughly five million US homes get re-roofed every year, and Tesla's pitch is simple: you're spending $25,000 on a roof either way — spend it on one that pays you back. The conventional panel line is the volume play, and it's the only track that can physically reach 100 GW of annual output. Expect the Solar Roof to carry the brand and the panels to carry the gigawatts.

We see the same split at our own counter. Customers who need a roof love the integrated pitch; customers with a ten-year-old architectural shingle roof just want the cheapest warranted watt on a rack. Tesla's two-track model is the first serious attempt by a US manufacturer to serve both without cannibalizing either.

On panel efficiency and cost per watt alone, Tesla loses to LONGi, JA Solar, and Jinko every day of the week. Those companies produce more gigawatts in a quarter than Tesla has produced in its history. So why does the 100 GW plan matter? Because Tesla is not selling panels — it's selling an integrated energy stack no panel maker can match:

As time-of-use rates spread, interconnection rules tighten, and grid instability makes backup a standard homeowner expectation, storage attachment rates on residential solar have climbed past 30% in California and are rising nationally. Every point of attachment rate is a point where Tesla's bundle beats a commodity panel plus third-party battery on simplicity, warranty, and commissioning labor. The 100 GW solar target and the ~300 GWh storage target are the same plan — neither works without the other.

Master Plan Part 3 puts the destination in one number: solar-plus-storage energy delivered below $0.05/kWh over system life — cheaper than any fossil alternative in almost every market on earth. Getting there requires four cost layers to fall together:

• Manufacturing scale. The solar industry's learning curve has held for four decades: every doubling of cumulative capacity cuts module cost roughly 20–30%. A 100 GW/year run rate is what the steep part of that curve looks like from the inside.
• Cell efficiency. Higher-efficiency cell architectures (heterojunction, and eventually perovskite-silicon tandems) squeeze more watts from the same glass, aluminum, and labor — dropping cost per watt without touching the BOM.
• Installation labor. Hardware is under half the cost of a residential system in the US; soft costs and labor are the rest. Tesla's standardized mounting and one-day-install targets attack the expensive half of the invoice, not the cheap half.
• Financing. As installed fleets mature and performance data accumulates, lender risk premiums compress — and at utility scale, a point of WACC is worth more than a point of module efficiency.

Here's the pricing context from the trenches: we've quoted Tier 1 410–450W modules through three full price cycles, and the trend line is brutal. Container-price panels that moved at $0.42/watt in early 2023 were crossing our dock under $0.18/watt by late 2025. If Tesla adds even 30–40 GW of real capacity by 2028, that curve steepens again. Great for project economics — painful for anyone holding inventory bought at last year's number.

The technology is not the bottleneck. The materials are. Producing 100 GW of modules a year consumes inputs at volumes that stress entire commodity markets, and each one has its own choke point:

Benchmark the ambition against the companies that actually own the module market today:

Read that table carefully and the strategy snaps into focus. Tesla will never out-ship Jinko in commodity modules, and it doesn't need to. Its 100 GW matters because every gigawatt arrives pre-attached to the industry's most mature storage and software stack — the thing the Chinese giants structurally cannot replicate in the US residential market.

Gigawatt numbers get thrown around until they stop meaning anything, so let's convert the target into units anyone in this trade can feel. Assume a blended average module wattage of 450W — the current sweet spot for residential and light commercial product. One hundred gigawatts a year is 100 billion watts, which works out to roughly 222 million modules a year, or about 608,000 modules rolling off the line every single day of the year.

Twelve million rooftop systems a year is more than double the total number of US homes that have ever gone solar. Nine hundred shipping containers a day is a port logistics operation on the scale of a small automaker. That's what the target means physically — and it's why the supply chain section of this article isn't filler. Glass, silver, frames, and freight all have to exist at that scale before a single panel ships.

From a distribution seat, the part of that math that keeps me up is the channel implication. Two hundred twenty-two million modules a year cannot move through traditional distribution alone; Tesla will push most of it direct. What remains for the channel is everything around the module — storage, racking, BOS, and the service layer. That's not a threat so much as a map of where the margin is moving.

Piecing together Tesla's public statements, permit filings, and analyst tracking, the ramp looks roughly like this:

Whether the number lands in 2028 or 2031, the direction is set, and the direction has consequences for every seat at the table:

One thing we tell contractors every week: don't wait for Tesla's ramp to change your business model — the price pressure is already here. The installers thriving right now are the ones who stopped marking up modules and started selling design, storage, and monitoring. Tesla at 100 GW only accelerates a transition that's already underway.