The useful thing about a shopping mall roof is that nobody expects it to be glamorous. It is large, expensive to maintain, and usually occupied by machinery. In 2018, the roof of Hanford Mall in California acquired a different job: hold 1.62 megawatts of Sunpreme solar panels. The developer projected nearly three million kilowatt-hours of annual production. That estimate belongs to the project team, but the reason for choosing the panels was plain enough. There was only so much roof, and the developer wanted more electricity from it.
Sunpreme's answer began with a mildly embarrassing observation about conventional solar panels. Their backs do very little. A panel can sit above a bright roof, pavement or pale ground while reflected light passes beneath it. Sunpreme designed cells and double-glass modules that could collect some of that light from behind. The concept has an almost comic neatness: if you own both sides of the glass, perhaps both sides should work.
- Sunpreme sold bifacial, double-glass solar modules for rooftops, carports and utility sites.
- Its commercial pitch was energy yield within a site's space and cost limits, not a cell-efficiency number alone.
- It developed technology in Sunnyvale and scaled manufacturing in Jiaxing, China.
- Its most useful lesson: a clever module still depends on the surface below it, the partner installing it and the rules governing its journey.
The second side has conditions
Bifacial sounds like a guaranteed bonus. It is not. The back of a panel needs a view of reflected light. A dark roof pressed close to glass offers less opportunity than a bright surface with generous clearance. Rows can shade one another. Racking can block light. Climate, dirt and the hour of the day matter. The payoff is measured in actual kilowatt-hours over time, not merely in the front-side wattage printed on a label. Sunpreme's technical idea was therefore inseparable from the site design.
The company called its cell platform Hybrid Cell Technology, or HCT, and used thin-film layers with a silicon wafer. Its Maxima GxB line put those cells between two sheets of tempered glass in a frameless module. In 2014 it introduced 280-watt and 340-watt versions. A 2015 smart version paired the bifacial module with a Tigo optimizer for panel-level monitoring, longer strings and rapid shutdown. A later HxB-400 product was aimed at commercial and industrial systems. These are products for people who design and buy electrical assets, not an app for people browsing rooftops on a Saturday.
There were adjacent advantages in the design. Glass on both sides gave the back an appearance a carport owner might tolerate overhead. Frameless construction could simplify some grounding arrangements. Datasheets described resistance to heavy snow and wind. But every such feature carries its own engineering bill: weight, mounting hardware, handling and the precise requirements of a particular roof. The customer buys a system, even when the invoice begins with a module.
A lab idea meets a very large invoice
Ashok Sinha, a former senior executive at Applied Materials, founded Sunpreme and led it as chairman and chief executive. The company's public profiles place its base in Sunnyvale, the familiar address for an unfamiliar wager. Its manufacturing story was less local. Sinha knew the solar supply chain was gathering in China. In 2009, a contact from Jiaxing suggested locating a factory there. Negotiations followed, and a production site came together with a speed Sinha later described to Fortune as extraordinary. Local incentives were worth about $5 million by his estimate.
The International Finance Corporation's 2011 disclosure reveals the scale of the bet. Sunpreme proposed an initial 30-megawatt manufacturing capacity in Jiaxing and an expansion project costing about $60 million through 2012. That figure was a project estimate, not a published receipt for money spent. It covered integrated production facilities, equipment, upgrades, working capital and research. IFC considered an investment of up to $25 million and records an investment that July. Company data separately reports a $50 million Series B round, but the public project document should not be mistaken for proof of that round's final size.
That is the peculiar asymmetry of climate hardware. The key idea can fit in a diagram. The route to a reliable, repeatable product requires capital equipment, high manufacturing yield, inventories, distributors, certification and buyers willing to underwrite decades of performance. Sunpreme described its factory as running around the clock with high yield. Those were company claims, but the capital plan itself makes the industrial burden tangible.
The customer buys a roof, not a record
Sunpreme's customer list is easier to understand as a chain than as a roster. Developers and installers chose the module; landlords, tenants and utilities received the electricity. At Hanford Mall, Blue Sky Utility developed the installation with Passco Companies, the property investor. Blue Sky said the available roof area, schedule and cost requirements drove its selection. It was the third project the developer had done with Sunpreme. That repeat purchase is more revealing than a laboratory efficiency chart.
“There was only one panel design capable of meeting the energy production requirements within the constraints of available rooftop space.”Ran Bujanover, Blue Sky Utility, on Hanford Mall
The company's other reported installations show the range of the pitch. In 2016 it said it had supplied 12.8 megawatts of GxB modules to an eastern US utility project. In 2019 a Bay Area shopping center completed a 750-kilowatt mix of carport and rooftop arrays using Sunpreme GxB380 smart modules with Tigo optimizers. Greentech Renewables showed the same technology in smaller pergola and awning projects. A mall roof, a parking canopy and a garden structure do not share a floor plan. They do share a question: how much energy can this specific piece of real estate make?
Partnerships widened the answer. Tigo added module-level electronics and monitoring. JDS Power installed and distributed smart modules. Greentech Renewables carried the product into its installer network. In 2018 Sunpreme announced an alliance with Honeywell Process Solutions to combine modules with solar plant management hardware and software for commercial and industrial sites. That was a bid to make output visible and manageable, not merely to make the glass more efficient.
The border was another component
There was a less photogenic test. Sunpreme's modules were made in China and imported into the United States. The company argued that its hybrid thin-film architecture should fall outside certain US orders covering crystalline-silicon photovoltaic cells. Customs and Commerce disagreed. In 2020, a federal appeals court upheld the conclusion that the silicon wafer in Sunpreme's module actively contributed to electricity generation and that the modules fell within the orders. The company's technical distinction did not win the legal distinction it sought.
That ruling does not tell us whether bifacial panels worked on a roof. It tells us that a manufacturing strategy and a product definition can collide at the border. In 2015, the reason to manufacture in Jiaxing looked persuasive: speed, supply chains and financing. The tariff dispute showed a cost that could not be read from the module datasheet. For any hardware company sourcing abroad, customs classification belongs in the model as surely as glass thickness and conversion efficiency do.
Sunpreme's later public record is thinner than its burst of announcements from 2014 through 2019. That limits any claim about current production, warranty service or market share. Its case remains useful because the underlying lesson travels well. Start with a resource others ignore. Prove it on a site where the geometry helps. Sell through partners who already own the customer relationship. Then price the factory, the installation and the border crossing before declaring victory. The back of the panel may be free real estate. The rest never was.