Thank you, Sharelynn. It's great you were able to share the progress that the engineering product management and commercial teams are making on our growth initiatives. This work is positioning Bloom for future growth. And most folks don't have the opportunity to see the advancement the teams are making across these initiatives. Also the point you and K.R. both made about what's unique to the Bloom technology in its flexibility to address multiple market opportunities by leveraging the same platform is important not only in our time to market, but also the flexibility of our investments.
Our energy server can support our expansion initiatives with minimal customization. Sure, our server will be optimized for each application. But it's the same basic design of a solid oxide fuel cell that we utilize the same supply chain, manufacturing capacity and engineering teams. This flexibility dramatically changes the economics of new product and market expansion opportunities. It gives us a cost and time to market advantage when entering these high growth adjacencies like marine, carbon capture in the hydrogen economy. Within our core, the growing emphasis and awareness on climate issues is accelerating the demand for clean, efficient natural gas fuel cell deployment for resilient baseload power and our cost down efforts continue to open up markets in the U.S. and internationally.
As Sharelynn described, we are building the direct in partnership origination models to access additional state in countries to capture this accelerating demand.
As the world moves forward with hydrogen, we're confident our solutions will be competitive in their inherent efficiency advantages in leveraging our current supply chain and manufacturing scale. If hydrogen adoption accelerates in markets and applications, we can allocate manufacturing capacity to meet the demand for electrolyzers in hydrogen fuel cells. And as we made progress in adding our solid oxide fuel cell for marine applications, it will come from the same manufacturing sites, and global suppliers for our land based servers. This singular platform creates tremendous flexibility. I'll spend more time on our capacity builds later. But I thought it was important to reinforce how key this is to our investment thesis.
Now let me spend some time reviewing the first quarter financial performance. We've kept the format for earnings release and supplemental information some of the last quarter. I’ll be referring the slide presentation posted on our website. We're off to a good start for the year. We had strong growth versus the prior year, and our mix of accepted deals resulted in attractive margins. We achieved a record first quarter acceptances of $359 million up 40.2% versus the first quarter of 2020. Revenue totaled $194 million up 23.8% versus the first quarter of 2020. Non-GAAP gross margins of $29.7 million increased 13.5 points in positive non-GAAP operating income of $2.8 million increased $26.2 million both versus prior year.
As you can see from the revenue and margin analysis, Slide 4 of the presentation, there were several factors underlying this performance.
First, based on our strong growth and acceptances, our product revenue was up 38.5% versus the prior year. Product margin improved 7.9 points as product costs were down 12% versus the prior year. Overall revenue growth was impacted, as many of these acceptances did not have installations. These are because the installation was done by our partner in Korea, or for a specific customer. The installation will be performed later in the year, thus reducing our install revenue 83.7% versus the first quarter of 2020. It's important to remember most installs are targeted at breakeven.
So as we find more installation partners, it will provide the margin improvement that we saw this quarter.
We are having discussions with potential partners that could provide the installation services and earn the revenue on future projects. Well this may reduce our revenue it should be a benefit to our margins and reduce our operating complexity. What may be most important in the margin analysis is that the service business reported non-GAAP gross profit in the first quarter of $1 million.
As we discussed during our service business presentation a few months ago, the significant improvements in our power module life cost reductions, and our actions to proactively manage the fleet have accelerated our expectations on profitability for our service business. I'd like to congratulate Glen Griffiths, [indiscernible] on achieving profitability in the first quarter and relay their goal to be profitable each quarter as we work towards our targeted services non-GAAP gross margin of 20% by 2025. We delivered adjusted EBITDA of $16.1 million for the first quarter, an improvement of $25.9 million versus the same period prior year in our adjusted EPS improvement of $0.27 versus the prior year. With respect to our cash flow and debt analysis on Slide 5, our usage of $89 million in operating cash or CFOA reflects an increase in working capital to support shipments later this year, without the benefits of receiving deposits from our customer financing vehicles. At the end of last year, we did not renew our 2020 PPA financing providers as they were tax equity constraint. We've already engaged with several new and former finance seers to support conversion of our backlog and growth.
Our first quarter did not require these new financing vehicles to be in place to facilitate acceptances. But not having them did reduce our CFOA just over $20 million, as we were not collecting the milestone payments we generally receive prior to acceptance. In the second quarter, we expect to close these financings to support the second quarter in nearly all of our 2021 U.S. C&I acceptances.
Before I leave the debt analysis, I want to note we adopted new accounting guidance for 2.5% green convertible notes due August 2025. Under prior guidance, we were required to allocate the principle between debt and equity because of the indebted conversion features. Beginning in the first quarter, we will be accounting for the entire unamortized balances debt.
While this increases the debt balance on our balance sheet, it does not change the amount we owe on the notes and we feel it more closely aligns with our capital structure. There is an impact to interest expense, as it will no longer amortize in the debt discount. And the first quarter results best reflect our current quarterly interest expense runway. Non-GAAP operating expenses of $54.8 million have increased $6 million versus the first quarter of last year. The increase is driven by investments in our front end originations capability both in the United States and as we announced a few weeks ago, by building our international resources under Azeez Mohammed.
We also continue to invest in R&D capabilities to support the technology roadmap. I would expect as we continue to invest in these areas, there'll be increasing operating expenses in the coming quarters.
While we only provide booking and backlog annually, in our early stage pipeline, we are seeing an increase in commercial momentum for our current product offering.
We are always on energy server offering 24/7 resilient baseload power with a greater percentage of large megawatt opportunities in that pipeline. The pipeline reflects the changes we've made to include new states, international and focus on larger transactions.
We are encouraged by this progress and look forward to moving these and additional opportunities through the pipeline and into bookings.
Our supply chain team continues to execute in a challenging environment. Even with the significant impacts of COVID, especially in India, and the logistic disruptions from the Suez Canal blockage, the team has successfully secured the inventories we need to run our factories.
We are working with our suppliers and logistics providers to proactively identify potential disruptions and mitigate. We're experiencing some increased costs for expediting, air freight and safety stock.
We are mitigating most of these increases through logistics optimization, but we still expect cost increases in the range of 1% to 2% of material costs.
We expect these costs to be temporary, but necessary to meet our customer commitments. Overall, we are pleased with our performance in the first quarter.
As I said at the end of last year, we believe these financial and operating accomplishments provide meaningful proof points on our growth journey.
Now, let me turn to manufacturing capacity investment. It's important for me to point out that we currently manufacturer our fuel cell stacks in California, and do the remaining manufacturing and final assembly in our plant in Newark, Delaware.
Our facilities in Delaware were constructed to support over one gigawatt of annual production. And we can scale quickly by adding additional manufacturing employees in that attractive market.
In fact, with minimal capital investment in the right labor planning, we can increase our capacity to nearly two gigawatts of annual production there.
In addition to Delaware, as our Korean volumes continue to grow, we are building capability in our joint venture with SK E&C to support not only that market, but also to help us enter additional markets across Asia.
Our pressing constraint is unlocking additional capacity is production in our fuel cell stacks at our Sunnyvale, California facility. We currently have about 200 megawatts of Bloom 5.0 revenue stack manufacturing capacity there, with no remaining additional space to increase capacity at this site. To support our growth expectations, and with the introduction of Bloom 7.5 this quarter, we secured 154,000 square foot facility for additional stack manufacturing capacity. The facilities in Fremont, California, close to our engineering team, which we think is important since this coincides with our introduction of the next generation of technology Bloom 7.5.
We are ordering the tooling required to build this platform.
As I discussed previously, we're investing $50 million to $75 million to bring additional 200 megawatts of capacity online over the course of the next year as we operationalize manufacturing a Bloom 7.5. The facility we've secured is large enough that for a total investment of $200 million, we can increase the capacity roughly to around one gigawatt of Bloom 7.5 fuel stacks.
Our introduction of Bloom 7.5 remains on track.
We are seeing the performance that we expected and we'll continue to introduce that product into additional field sites.
We will cadence Bloom 7.5 manufacturing investment for the demand for additional capacity. We believe this to be an attractive return profile to allocate capital.
We have the additional confidence given the stack capacity to be utilized across applications.
Given the inherent advantages that we have in our platform, we believe that our investment strategy, which is to focus on technology innovation of the platform, manufacturing excellence and the distribution of our product in our core and adjacent markets, with meaningful partnerships is the most efficient value maximizing approach. From an investment thesis, the diversification across end markets provides us with significant opportunity to de-risk our investments. That's what gives us such strong confidence to make the investments in technology and additional manufacturing capacity. On Slide 7, we highlight our 2021 outlook. And we're reaffirming all of our 2021 targets.
For revenue, after a strong start, we maintain our expectation to be between $950 million to $1 billion based on our mix of expected acceptances and increases in installations over the next several quarters.
We expect total year non-GAAP gross margins to be around 25% and positive non-GAAP operating income of roughly 3% of revenue.
As I previously mentioned, we expect to continue to invest in our sales and marketing and our research and development increasing each quarter. And I would expect operating expenses as a percentage of revenue for 2021 to be similar to 2020. We're also maintaining our outlook on CFOA as approaching positive.
While we do not provide quarterly guidance, there are a few aspects of the second quarter that I wanted to highlight.
All of these are accounted for in our 2021 framework, and do not change our yearly targets. On revenue, I would expect second quarter revenue growth to be similar to our annually targeted revenue growth, with installation revenue more in line with the 2020 as a percentage of revenue. Gross margins may be sequentially lower, due to one large transaction in the quarter that was booked 18 months ago, and has lower product margin than our norm as well as the dilution from increased installations. It's likely that when incorporating second quarter into the first half 2021 results, non-GAAP gross margin percentages will be in line with our total year expectations. These factors combined with an increase in operating expense will result in sequentially lower operating income in the second quarter, that are all factored into a full year targets. In summary, we have a strong operating performance in the first quarter and our very confident of our future. We're gaining momentum in our commercial operations and we're seeing opportunities with new customers in new geographies.
We are investing in our technology, manufacturing and front end originations teams.
Our service business has improved is demonstrating in its results. And we are reaffirming our 2021 framework. We feel Bloom Energy is well positioned and has the product team to be the leader in distributed generation. And we're doing this with a real focus on discipline and financial management, and operational excellence, which creates value for our shareholders. With that operator, let's open up the line for questions.