SoloPower said in late September 2012 that it had started up its large factory in Portland, Oregon. SoloPower then-CEO Tim Harris told us at the time that he expected the company to produce 20 MW – maybe 30 MW — of solar panels per year by the end of 2012 and ship 2 to 5MW of solar panels during the fourth quarter. In SoloPower’s confirmation of the layoffs this week, the company says it “will begin commercial shipments to customers this month.”

But according to The Oregonian, SoloPower is looking to sell millions of equipment from its headquarters in San Jose, Calif., it’s seen the departure of high-level executives like the CEO, the President and the CTO in recent weeks, and the managing director of SoloPower’s lead investor Hudson Clean Energy Partners recently resigned. These are all worrisome signs for a startup that has raised over $200 million in funding. The factory originally was supposed to be 400 MW factory and cost $350 million.

SoloPower was awarded a $197 million federal loan guarantee to help it build out the factory. That loan is supposed to be able to be drawn down on when it has its first production line up at the factory. That’s the same program that funded Solyndra. SoloPower has drawn down on a state loan according to the Oregonian.

This article was updated at 11:48AM PST on March 5, 2013, to clarify the status of its loans.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• After Solyndra: analyzing the solar industry
• Flash analysis: the Fisker debacle and its implications on investing, innovation, and government incentives
• The fourth quarter of 2012 in cleantech

New investor Crosslink Capital led this E Series round, joining previous backers Battery Ventures, BlueRun Ventures, El Dorado Ventures and Mohr Davidow Ventures. The cash influx brings total venture funding to $41. 5 million to date for the six-year old company.

Coupa, based in San Mateo, Calif. competes with companies like Ariba which started out in on-premises software but is adding more cloud services.

Business customers use these e-procurement services to track contracts, purchases and suppliers. In a statement, Coupa CEO Rob Bernshteyn said the company has closed 13 consecutive quarters of growth and has 200 customers in 40 countries.

In e-procurement — as in other vertical application areas — traditional software vendors like IBM, SAP and Oracle are building (or buying) cloud-based services to compete with pure play cloud players like Couda. Other e-procurement competitors include JDA, BravoSolution, and Emptoris, which IBM bought in December.

Photo courtesy of Flickr user Images_of_Mone

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• A near-term outlook for big data
• Sector RoadMap: SQL-on-Hadoop platforms in 2013
• Big data 2013: key trends and companies to watch

Affine Systems' object detection

A few days ago, we reported that Affine Systems had raised a $5 million round of financing. Well, there was one detail in the SEC filing announcing the round that we missed on first glance: Michael Mathieu, former CEO of video ad network YuMe, has taken the chief executive role at Affine.

Mathieu’s experience is primarily in sales, which could prove beneficial as Affine attempts to get more clients for its semantic video ad technology. The exec spent three years as chief executive of YuMe before Co-Founder Jayant Kadambi took back the reins. Prior to his chief executive stint at YuMe, he served as president of the Internet Division of Freedom Communications, and before that, was an SVP at United Online.

While co-founder and former CEO Michael Sullivan will hand over the chief executive title at Affine Systems, he’s not leaving the company and is still listed as an executive in the filing. (We’ve reached out to Affine to find out exactly what his new title is.) After hearing back from Affine, we’ve learned that Sullivan will take over the CTO role. With the latest round, Affine has raised $13 million since being founded in 2007, with investors including Highland Capital Partners and Crosslink Capital. Clients and partners include Adap.tv, HBO, Sony, TubeMogul and Microsoft’s Xbox 360.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Where the next-generation console fits in today’s video game market
• How consumer media will change in 2013
• Controversy, courtrooms and the cloud in Q1

SeaMicro, which recently expanded to a new 68,000-square-foot headquarters in Sunnyvale, Calif., has designed server technology that consumes a quarter of the power of a regular server but packs more than 2,000 CPU cores and costs about $139,000. The innovation uses Atom chips and its own specially designed silicon to manage the networking. The bet is that Internet giants will be willing to buy servers from a young startup to save on the ever-increasing energy costs of computing.

SeaMicro’s technology, which it launched about a year ago, re-architects server components to consist of hundreds of low-power processors and a box that consists of 10 racks and consumes less than 2 kilowatts of power. The server itself can fit on a credit card and eight of them fit on a single motherboard. The basic idea is to strip all the extra hardware and functionality out of the servers to provide enough performance, but not deliver overkill, leading to a lot less energy consumed. Most web servers just need to offer up a website or grab a photo and don’t need to deliver super high performance that gulps power.

One indicator that SeaMicro’s tech could be huge: Intel manufactured a dual-core Atom chip to go inside SeaMicro’s box, and it was perhaps the first time that Intel built a chip especially for a startup. Intel has been trying to defend its territory against low-power ARM chips moving in, and SeaMicro launched the second iteration of its server tech in February. SeaMicro also counts Skype, France Telecom and Mozilla as customers.

Other startups like Calxeda are also working on low-power server technology. As Internet companies expand their data centers and devices get always-on connections, the amount of energy that is going to power servers in data centers will continue to grow. Reducing that cost is becoming a competitive advantage.

SeaMicro was previously backed by Crosslink Capital, Khosla Ventures and Draper Fisher Jurvetson, and those investors are again listed on this latest filing. The Department of Energy also awarded SeaMicro a $9.3 million grant last year to help it field test its servers, and that was one of the largest awards allocated out of $47 million in funding for 14 data center efficiency projects.

SeaMicro was one of our 10 Big Ideas companies at Green:Net and also a Structure50 company. Come hear more from SeaMicro CEO Andrew Feldman at our Structure event on June 22 and 23 in San Francisco.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• The fourth quarter of 2012 in cleantech
• Cloud computing’s impact on chip and hardware design
• Cloud computing infrastructure: 2012 and beyond

This is Mavenir’s fifth round of funding. The round — led by August Capital and including existing investors Alloy Ventures, Austin Ventures, North Bridge Venture Partners and Greenspring Associates — brings Mavenir’s total funding to $105 million since 2005, a phenomenal amount when compared with the much smaller rounds raised by other companies offering IP-based communications.

However, unlike services such as Skype, Fring, Twilio and others that seek to deliver IP communications over the top of existing broadband networks, Mavenir wants to throw operators a lifeline to help prevent them from becoming a dumb pipe. Mavenir’s gear sits inside an operator’s network and lets it offer IP communications services like video calling, more accurate presence and other ways to make a seamless transition from the type of collaboration tools offered on wireline networks to mobile ones. The company is already working with three U.S. operators, including an announced deal with Metro PCS. The new funds will help Mavenir expand its footprint especially, now that operators are deploying faster 4G networks that can truly deliver on the promise of mobile collaboration.

During a lunch two weeks ago with Shubh Agarwal, Mavenir’s VP of marketing, he showed off some nice demonstrations of the services Mavenir Networks can provide. The video calling was especially compelling because it could run over the operator’s network or the Wi-Fi network. And by using the operator’s network, the quality was assured. If the operator’s network was unable to make the call, it didn’t offer video as an option, but would connect voice: a graceful degradation of service that’s far better than just being cut off mid-call or being unable to make a call at all.

The services also offered some nifty blocking features and a deeper granularity around availability. For example, I could be available for video calls only to my family, while on weekends, my coworkers could see me online only for voice and texts and PR people wouldn’t see me online at all. Unfortunately, Mavenir’s more granular features could end up dead on arrival if carriers don’t price them appropriately for consumer adoption — a problem in an industry where Verizon Wireless charges $10 a month for a navigation service I can get from Google for free.

Mavenir gets paid either way, since carriers have already bought the gear, but Agarwal is confident that carriers have learned from their past battling free or cheaper over the top services. He suggests that with LTE networks, the carrier has a level of control that will boost the quality and dependability of such a service and carriers will price the offerings competitively. Of course, I’ve heard this before from other gear companies and even operators, but the tidal wave of adoption is still driven by cheaper or free services.

Still, Mavenir continues to raise money, boost its offerings through acquisitions or organic efforts and sell gear to operators. Perhaps with the advent of 4G networks, operators will have a chance to change the narrative from one of good-enough, free services to high-quality, paid ones.

Disclosure: Mavenir Systems is backed by Alloy Ventures, which also backs GigaOmni Media, the parent company of GigaOM. Alloy’s Ammar Hanafi is on the board of both companies.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• CES 2012: a recap and analysis
• 12 tech leaders’ resolutions for 2012
• LTE-Advanced: what it is and isn’t

At the same time, Hawaiian solar concentrating startup Sopogy is also looking to raise $30 million, according to a filing also revealed Wednesday morning, and has closed on $4.5 million of that round. Yesterday, the CEO of solar panel startup Stion told us the company plans to raise between $100 million and $150 million in an IPO sometime in 2012.

Solar gear companies still seem able to raise funds, despite many predictions that 2011 will be the year solar investors will turn to low-capital investments like software and services.

San Jose, Calif.-based SoloPower has developed a flexible solar panel, which received UL certification in September: a certification required for installation in many U.S. regions. Its panels are made of copper-indium-gallium-selenide cells, which represent the next generation of thin-film solar technology.

The majority of the solar panels on the market use crystalline silicon solar cells, which are fragile and rely on glass to protect them. Thin-film solar panels, on the other hand, contain ultra-thin layers of alternative versions of silicon or other semiconductors and don’t necessarily need glass. However, the world’s largest thin-film maker, First Solar, sandwiches its cadmium-telluride solar cells in between glass.

Producing these cutting-edge solar panels at scale will require hundreds of millions of dollars. SoloPower raised nearly $45 million in debt financing about a year ago, according to a regulatory filing. Close to half was paid to co-founders Bulent Basol and Homayoun Talieh, in what appeared to be the last chapter of a lawsuit settled between Talieh and SoloPower.

Back in 2008, Venture Wire reported that SoloPower had raised a grand ol’ $200 million to scale manufacturing up to a 100-megawatt-per-year plant (via VentureBeat), though Solopower never confirmed that round size with me.

In late 2009, SoloPower said it was looking for a $190 million loan guarantee from the Department of Energy to build a high-volume manufacturing plant, and would raise more funds to meet the equity-share requirements of loan guarantee. To date, that loan guarantee hasn’t come through.

SoloPower is also now reportedly seeking a $20 million loan through Oregon’s State Energy Loan Program for a 75-megawatt capacity plant in Wilsonville, Ore.

Related reports on GigaOM Pro (subscription required):

• Getting Solar Onto the Smart Grid
• Renewable Energy Charging Up Electrical Transmission Tech
• New Opportunities in the Smart Grid

For the details on how the technology works, I turned to broadband infrastructure guru Stacey Higginbotham, GigaOM Senior Writer, to give us an explanation. Stacey says SeaMicro has built a box that contains 512 Atom CPUs, a petabyte of storage, and costs less than $100,000. The company hopes to use that box to sell into a market which chip-makers have largely ignored: There’s a huge range between low-power mobile chips and incredibly speedy high-performance chips, but most chip makers are neglecting opportunities in the middle range.

Stacey says using high performance chips in servers for regular old web page serving is like “using a nuclear bomb to take out a car.” And the excess processing power wastes a whole lot of energy. Since chip makers have been ignoring computing power needs in the mid-range, SeaMicro decided to re-architect server components and use the DOE funds to field test servers consisting of hundreds of low-power processors. SeaMicro explains the innovation as:

By efficient use of tiny interconnected Central Processing Units (CPUs) within a single server, demonstration of this patented technology is expected to save 75 percent of the computing energy over conventional servers.

If every regular old data center server used this technology, it would save a shocking amount of energy — and money on web companies’ monthly energy bills. And SeaMicro’s technology could remake computing power by just thinking outside the box and using already-available tools and components. Other companies looking at similar issues include Austin, Texas-based Smooth Stone.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• The fourth quarter of 2012 in cleantech
• Cleantech and investment in 2013
• Financing the next generation of great cleantech ideas

SoloPower also announced this month that its founding chief executive, Homayoun Talieh, is leaving and that it has selected Lou DiNardo, a partner with SoloPower investor Crosslink Capital, as its interim CEO. The news came just a few months after the departure of Rommel Noufi, who had been the company’s vice president of research.

At the Intersolar North America conference in San Francisco last week, we sat down with Alain Harrus, a partner at Crosslink Capital and a SoloPower board member, to get an update on what’s happening at SoloPower and hear his thoughts on some of the changes the industry is seeing in solar investing. Here are excerpts from our conversation:

Q: So what really happened with the CEO? Why did SoloPower decide to make the change?

A: [Changing CEOs] happens very often and it’s kind of the natural evolution of these companies, especially when you’re starting from nothing. This started with two people; it really was just an idea. [When a company matures and enters mass production, you often] need a different set of skills. A good analogy is a triathlon or a relay race. You need different people who have different capabilities.

Q: The company put an interim CEO in place, rather than waiting to find a permanent replacement. Why did the CEO leave before you’d found a permanent replacement?

A: We selected an interim CEO just because of timing. Sometimes it takes longer to get a new CEO — it might take 90 days, [especially because] solar is in a downturn. [We decided to change CEOs right away] in order not to lose momentum. We don’t want to miss the market window. It’s a matter of getting the right staff to get to high-volume manufacturing.

Q: Why did Rommel Noufi also leave the company?

A: Rommel was on a leave of absence from NREL and was commuting from Colorado. He had been here for a year, and it was very interesting, but he had been at NREL for 27 years and all his family was there [in Colorado]. His decision was driven more by personal reasons.

Q: Why do you think it makes sense to pursue this market – and a new factory – when supply outweighs demand and SoloPower faces a crowded field of thin-film competitors?

A: Right now, solar is in a downturn, but all the companies in CIGS are just making miniscule amounts of product, in the tens of megawatts. The ramp for product is 2011 or 2012. Then, the demand will be a lot larger because of what I call “the sleeper markets,” the United States and China, will wake up, and we’ll also see strong markets [elsewhere]. By the time we get to the 8-10 MW levels, it will be a pretty good market for CIGS.

Q: How do you expect to be able to compete with First Solar and other competitors?

A: The market is so large, no one person can meet all the demand. The right price point and efficiency, and you have a business. So it’s not about beating First Solar; it’s about being as close to First Solar as possible. As long as you’re in that zone [of cheaper panels], you can sell. And Solopower wants to be a flexible panel, not flat-plate glass [like First Solar]. You’re going to see more product differentiation going forward.

Q: Venture-capital investments in solar have declined. Is the VC appetite for solar becoming saturated?

A: There’s still a lot of interest in the next generation of photo absorbers — not just concentrators, but also multijunction and III-V materials — in better system monitoring and management and in the [energy] storage area. These areas are still very active, with the bar a little higher and a lower level of investment. [Reporter's note: Multijunction cells have multiple layers of semiconductor materials to boost their efficiencies, while III-V materials, such as gallium arsenide, also have the potential to reach much higher efficiencies than cells commonly used today.]

Q: Valuations for solar companies are down significantly. Do you think they will come back to the levels we were seeing most of last year, or do you think those valuations represented a spike and were too high?

A: Over a long period of time, valuations will go up. As long as companies are creating earnings and value, the markets we’re talking about are really large –- trillion-dollar markets. There are many areas where low-carbon energy could have a transformative impact.

Q: Some VCs have expressed concern that the government has begun acting as a VC, but may not be best-equipped to pick the winners. What do you think?

A: I think the money the government is spending is extremely well-spent money. Some will be conventionally spent -– such as in tax credits -– and some is going to be, frankly, VC money. But if you think about it, in some ways, VC creates enormous value. Even if [a venture-backed company] doesn’t succeed, the industry as a whole learns something and [the investment] creates jobs for people for two to three years, or the life of the company, and then they go on to start other companies. It’s not a return from a financial point of view [if the company fails], but it’s a different return. You could argue it provides a return by training or educating the workforce.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• After Solyndra: analyzing the solar industry
• The race for cost-effective and efficient solar power
• Warren Buffett and the true value of solar

Lou DiNardo, a partner with SoloPower investor Crosslink Capital and a semiconductor-manufacturing veteran with operating experience at companies such as Analog Devices, Linear Technology and Intersil, will take over as interim president and CEO.

San Jose, Calif.-based SoloPower, which is developing copper-indium-gallium-selenide cells and panels, gave no explanation for the changes other than to say it is “accelerating into its next stage of growth as it prepares products for market introduction and nears high-volume manufacturing.”

The problem for SoloPower, like many other thin-film companies, is that it’s currently in the difficult stage of trying to reach commercial-scale manufacturing; it’s planning to build a solar-panel factory and start high-volume production next year. To build that factory, SoloPower is betting on receiving a $190 million loan guarantee, and the company said in the release this morning that the U.S. Department of Energy has selected it “for due diligence and negotiations” for a loan guarantee. That sounds like it’s reached the technical and financial viability evaluation stage of the DOE loan program, which is laudable but can take a long time. (Just look at Tesla Motors, which waited for more than two years to get its guarantee.) We’re waiting to hear back from the company on this.

SoloPower didn’t give any details about the expected size or location of its plant. It has kept fairly quiet as a whole, with only three press releases — before this one — on its web site since 2007, when it raised $30 million in its second round of financing. The news about the loan guarantee and the factory is definitely good news for the company, but doesn’t explain why it has replaced its chief with an interim CEO, and whether or not this was part of a planned change as a result of expected growth.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• After Solyndra: analyzing the solar industry
• Ups and downs for cleantech in Q1
• Green IT Q1: Cleantech Breaking Out — and Bracing for Hard Times