Two routes give the answer, and they agree within about 10%. Specifically, the first multiplies rated watts by about 10 milligrams per watt. The second multiplies cell count by silver per cell. Both appear in the table, priced at around $55 an ounce, near the low end of October 2026 trading.

Products vary by design, and the value moves with the silver price. In addition, the gram figures are our arithmetic from published averages. Module sizes follow pv magazine's March 14, 2023 report on a 182 mm TOPCon series. It lists 108-half-cell panels at 415 to 435 watts and 144-half-cell panels at 560 to 580 watts.

Where Does the Silver Sit Inside a Solar Cell?

Silver does one job in a cell. It forms thin lines, called fingers. Specifically, they collect the current the silicon makes and carry it to wider strips, called busbars. Silver is the best electrical conductor among the metals. As a result, the lines can stay very thin without wasting power.

Most of the metal sits in those fingers. According to UNSW researchers in 2023, fingers held more than 75% of the silver in industrial screen-printed cells. Their measured finger widths ran from 22 micrometers on PERC cells to 35 micrometers on heterojunction cells. In fact, that is finer than a typical human hair. Therefore, shaving a few micrometers off each line saves silver across billions of cells.

What Is Silver Thrifting, and How Far Has It Gone?

Thrifting means using less silver per unit of power with the same design. Engineers print finer lines, add more busbars, and gain efficiency. As a result, each watt needs less metal. Substitution is different, because it would replace silver with copper.

In fact, the size of the change is large. According to UNSW researchers citing a 2022 peer-reviewed study, average silver use per watt fell about sixfold since 2010. It dropped from almost 60 milligrams to around 10 milligrams. Earlier ITRPV results, as cited by UNSW, showed TOPCon cells at 17 milligrams per watt. The 17th edition puts them near 10 for 2025.

Those two readings come from different ITRPV editions. For that reason, treat the gap as a direction, not a precise measurement. Even so, the direction is clear. Silver per watt fell for TOPCon, heterojunction and PERC cells between the editions. Moreover, the ITRPV roadmap expects further cuts. It projects TOPCon cells at roughly 57 to 60 milligrams per M10 cell by 2036.

Why Do Solar Manufacturers Thrift Silver?

Price drives it. In March 2026, pv magazine reported that silver paste accounted for more than 20% of total solar cell costs. In April 2026, it reported Silver Institute analysis that put paste at roughly 10% to 20%. The exact share depends on cell type and the silver price.

That share explains the engineering effort. At 10 milligrams per watt and $55 an ounce, silver costs about 1.8 US cents per watt of cell. The 60 milligrams per watt used in 2010 would cost about 10.6 cents at the same price. Makers compete on cents per watt. Therefore, every milligram saved lowers cost.

This is the price mechanism doing its job. A scarce input gets more expensive. So its users find ways to use less. Nobody ordered thrifting. Instead, the price signal produced it.

Why Do Two Counts of Solar's Silver Disagree?

Two respected sources count solar's silver differently, so you will see two different totals quoted. The Silver Institute and Metals Focus published the World Silver Survey 2026. It put solar silver demand at 186.6 million ounces in 2025, down 6%. It forecasts about 151 million ounces for 2026, down 19%.

According to pv magazine and PV Tech, the ITRPV counted about 7,244 metric tons of silver in 2025. That covered 706 gigawatts of shipped modules, or roughly 233 million ounces. PV Tech put it at 21.4% of world silver supply, against about 17% in the Survey.

Dividing each total by 706 gigawatts shows the difference. The ITRPV total implies about 10.3 milligrams per watt, matching its own cell average. The Survey total implies about 8.2. Different counting bases explain part of the gap. We cannot say how much. Both sources agree on direction: less silver per watt, and falling solar demand for the metal in 2026.

What Could Cut Solar's Silver Use Next?

Copper plating is the leading candidate. The Fraunhofer Institute for Solar Energy Systems ISE tested it in April 2026. Its team reported a plated process that cut silver to 1.1 milligrams per watt-peak. Notably, the cells were M10 TOPCon and reached 24% efficiency. That is about 89% below the 10 milligrams per watt industry average. Copper carries the current, nickel blocks copper from entering the silicon, and a thin silver cap protects against oxidation.

Treat it as a pilot-line result, not a production fact. According to Fraunhofer's Dr. Sven Kluska, the method could be established in the market within two to three years. However, Fraunhofer says makers would need to add plating equipment as an upfront cost. It also says printed copper for TOPCon cells is still in the testing phase.

Two outcomes are possible. Thrifting is likely to continue, because the roadmap expects further cuts. Meanwhile, substitution is possible too. Its timeline belongs to engineers and factory budgets.

What Does Thrifting Mean for Silver as Sound Money?

Silver does two jobs. First, it is an industrial input. Second, it is a monetary metal people have held for thousands of years. Thrifting trims the first demand pool. However, it does not touch the second.

Supply is the other half of the story. According to the Silver Institute's February 2026 outlook, primary silver mines supply only about 28% of mine output. The U.S. Geological Survey describes silver as mainly a byproduct of lead-zinc, copper and gold mines. For instance, a zinc mine expands when zinc pays, not when silver does. So makers can cut their silver use within a few years. Miners cannot add silver on the same schedule.

That asymmetry explains why the market can run a deficit while solar demand falls. According to the World Silver Survey 2026, the deficit was 40.3 million ounces in 2025. It forecasts 46.3 million for 2026. For the market-level argument, see our analysis of solar's silver cutback and the widening deficit. For the market-level argument, see our analysis of solar's silver cutback and the widening deficit. For the full balance of supply and demand, see the supply and demand guide.

Over a three-to-five-year horizon, silver per watt matters more than panel count. Watch whether plating moves from pilot lines to gigawatt factories. If you hold physical silver as a store of value, watch both signals. They show how much industrial demand you can count on.