The people
who made EVGA.

Chris Bencivenga ran our entire customer-service team. He was one of the most unflappable, unshakable people I have ever met.

Under Chris, customer service felt like an independent wing of EVGA, with its own confidence and an extraordinary ability to get things done.

Andrew was the emperor, but the customer was God.

Part of that came from a very aggressive escalation structure. If a problem was not being handled properly, it could reach Joe Darwin or Andrew rapidly. Nobody was eager to explain to Andrew why a customer was still waiting for something we should have solved. There was considerable pressure to put things right.

But there was also room to act. There were rules, and different levels of authority, yet in my experience there was remarkable discretion to consider a customer's circumstances and find a solution. An unusual case did not disappear into a rulebook. Someone could look at it, take responsibility, and decide which levers to pull.

That taught managers to be resourceful. Sometimes the solution required real creativity. You learned to ask what could actually be done for this person, with this problem, today.

Andrew genuinely believed in great service. Chris and the people around him made that belief work in practice. The independence, the pressure, and the trust all belonged to the same system. I have never encountered quite that combination anywhere else.

The EVGA forums were an enormous, beautiful thing at a time when that kind of place made sense on the internet. They gave the company a world around it: people, conversations, problems, discoveries, and an almost continuous sense that something was happening.

It was almost a little Matrix around EVGA. If you only remember the graphics cards, it is easy to forget how much life existed there.

Some forum moderators also worked as remote customer-service agents. They formed a remarkably experienced first line for customers, resolving problems themselves or knowing exactly where in the chain of command to take an issue for a quick solution. The connection between the forums and the support team made the distance between a public conversation and somebody doing something useful remarkably short.

I remember being astounded by how many calls those agents resolved before they ever reached our call center, and how consistently they could do that over long stretches. Their experience made a real difference to the customers and to the team in Brea.

Looking back, that was kind of insane.

You could go onto the forums, find somebody struggling with a problem, reach out directly, and help them solve it. If you went out of your way to do that, you could receive special recognition. That initiative was part of what being a good damn employee meant.

Other people on the team did this, too. I was one of the people encouraged by a culture that already valued reaching out and helping. I remember finding a customer's issue on the forums, calling them, and getting it sorted out. The customer left feedback indicating that the exchange had made their day. Jaeson Wilda personally congratulated me for providing this level of service.

I was elated.

Afterward, I started looking through the forums for other customer problems I might be able to resolve immediately. There was a very simple appeal to it. I knew something that could help. I could get in touch. Somebody who had been having a bad time in a forum might end up both surprised and happy to have their issue resolved with a direct call and then turn around and make sure everyone who passed through that forum thread knew it by posting online about how great the customer service was. This was part of the service we worked to provide to the community.

That was one of our principles, and the company rewarded it. The culture taught you that a customer's problem was worth noticing even before it arrived at your desk.

A support agent can promise all sorts of things. Somewhere, somebody has to make the next part happen.

The speed and consistency of the RMA department—the team handling returns and replacements—and the warehouse were a huge part of what made our service work. Those teams could turn something around quickly enough that the customer felt the company had actually moved for them.

The RMA department deserves a thank-you of its own. I remember their ingenuity in taking usable parts and “Frankensteining” them together into cards that, as far as I knew, had never existed as standard models. They were fully functional and made excellent RMA replacements.

To me, that was some of the most brilliant recycling and practical thinking I have encountered in any company. They could see what was still useful, understand what would work together, and turn those parts into something that could serve a customer again. That kind of resourcefulness deserves real credit.

Jayson Cane was a logistics master out in the warehouse. He kept things running, and he was one of the most wonderfully personable people you could ever hope to work with. Even when he seemed to have the world on his shoulders, he was still happy to help you with your issue.

That quality is easy to leave out when people write the history of a hardware company. It deserves to be remembered. A reputation for service depends on the people who answer the phone and the people who make good on what was said.

At its best, the whole operation had the assurance of a well-run Swiss hotel. People knew their work. They knew whom to ask. They cared whether the next part happened properly.

Like a hotel, our customer-service team had a rigid hierarchy, from level zero up to supervisor. That structure shaped rewards and bonuses, the kinds of problems each level handled, and the service that agents at that level could provide. People understood their responsibilities and where to take something that needed more authority. Honestly, I thought the way it all fit together was amazing.

That, to me, is how you turn a commodity device into a luxury product. Other companies could buy the same GPU. The experience of owning the finished card could still feel completely different. EVGA was doing for graphics cards what Lexus does for cars: making the care around the product part of the reason to choose it. The Swiss-hotel comparison is about that kind of service—competence and attention that you feel throughout the experience.

The Lexus comparison is my own. Its account of omotenashi describes anticipating customers' needs and treating owners as guests.

Some of the guys had spent decades working in customer service. They enjoyed the steadiness of it: an environment they understood, people they liked, and customers they could help. For me, it was a place to grow up. For others, it was a place where they had found a rhythm and a team worth staying with.

There were many more people than I can name here. I miss that team. They made an exceptionally demanding job feel like work we could do together.

Jaeson Wilda was my initial direct boss and trainer. He was also the hot-sauce and hot-pepper eater I mentioned in the first essay. His congratulations after that customer call stayed with me because I respected his judgment.

Before Jaeson trained me, Dave Biel had trained him. Dave had been his supervisor, although I couldn't give you the exact title. He left not long after I joined.

Dave wrote user guides, FAQs, and much of the technical material our customer-service team relied on. He also helped test and tune the SC17 laptop and work out how we would support it: batteries, RAM, cracked screens, and the problems customers would bring to us.

The Jedi-master comparison was definitely not lost on Jaeson: Dave had trained him, and now he was training me. Between their technical knowledge and their understanding of customers, I think those two could have run an incredible research lab.

I remember Andrew telling us how proud he was to have NVIDIA as a partner. That pride is part of what makes the ending so difficult for me to understand.

When I bought my first EVGA graphics card at about fifteen—the 8800 I mentioned in the first essay—I was so ignorant of how the industry worked that I thought EVGA was NVIDIA. That was how closely the two names were connected in my mind.

Years later, I understood the distinction. I also understood how much a partner could add: high-performance versions of NVIDIA's products, world-record attempts, a loyal community, and customer service that helped make people love owning a GeForce card.

Why couldn't that have become NVIDIA's enthusiast performance arm? Something with the place in people's imaginations that AMG has alongside Mercedes?

From my perspective, an acquisition by NVIDIA looks like an opportunity worth imagining. I do not know whether Andrew would have wanted to sell to anyone. I have no idea whether either company considered it, and I do not wish to imply that they did. I was never, ever high enough in the company to be in those conversations.

I am very glad of that. They would have given me panic attacks.

EVGA is deeply personal to Andrew. He cares about performance, quality, talent, and the experience of owning the product. Any future he could accept would, I think, have needed to respect that. Those seem to me like convictions NVIDIA could have made extraordinary use of.

EVGA's contribution began with relationships: distributors, system integrators, and retailers. Maintaining those relationships took time, trust, and people who knew how to get products sold, built into systems, and supported in the real world. That work took weight off NVIDIA, giving it more room to concentrate on the next great chip design. The expertise had value, and providing it had a human cost.

Alongside that work, EVGA was a specialist performance house. NVIDIA supplied the silicon; our people worked on how to power it, cool it, tune it, and support the person who bought it. That is what I mean by the AMG comparison: giving people a reason to want the specialist's version.

You could see it in the product range. There were reference designs, EVGA cooling such as ACX, and factory-overclocked cards. Moving through models such as SSC and FTW, you could get higher clock targets and more ambitious supporting hardware, including custom boards and power delivery designed to give overclockers more room to work. The combinations depended on the model and generation, but the progression toward specialist performance was clear.

Features such as selectable BIOSes, replaceable BIOS chips, and CPU-less flashing gave motherboard owners room to experiment and recover. Even the higher-end GPUs had multiple BIOSes and the ability to swap between them with a tiny toggle. We compared GPU BIOS tuning to working with a car's ECU: adjusting settings and limits to see what the hardware could do. That connection came naturally to the enthusiasts on our team who knew Southern California's early-2000s tuner scene.

I only knew the Brea side firsthand. The Taiwan team had engineering and repair expertise that deserves just as much credit. GamersNexus showed some of it in its visit during the GPU division's final days: after a card failed, the team moved its carefully selected, already-lapped GPU onto another board and got it working again. The computer equivalent of a heart transplant.

Andrew was an Enzo-like figure in that environment: fiercely attached to the brand, demanding about the work, and good at finding people who could meet those standards. There was a racing-team mentality to it. The records mattered, the finished product mattered, and so did the name on it.

He could be difficult to work with. I still believe the team he assembled and the standards he set were worth the effort of finding common ground.

Technical background: GamersNexus's examination of an FTW circuit board and EVGA's Z390 DARK manual (BIOS features on printed pages 23, 137–139, and 143). The automotive comparison is about the team's role and approach to the work.

There was another part of EVGA's identity that mattered to me: its relationship with the press.

We often worked with collaborators and influencers who lived in our area. We liked people who told the truth, because we wanted the truth about the products to be visible. For a performance brand like ours, a close examination could reveal the attention to detail, the thinking, and the decisions that made our work distinctive.

The closer you looked, the more there was to talk about. Why that circuit board? Why that cooling arrangement? What was the design trying to achieve? There was engineering, a vision, and a strategy around NVIDIA's silicon that belonged to EVGA. Other board partners made their own choices. We wanted people to understand ours.

To me, an independent reviewer willing to look carefully was valuable. That kind of scrutiny gave the work a chance to speak for itself.

But the relationship among EVGA, the press, and NVIDIA often felt unnecessarily tense. From where I stood, there was too much confrontation over how NVIDIA's brand was represented. I did not understand why it needed to be that way, and I still don't.

I cannot speak for NVIDIA's internal reasoning. What I remember from our side is the feeling that explaining what made an EVGA product special could become more contentious than it needed to be. I thought the care we put into those products, and the willingness to let people examine it, could strengthen the appeal of the NVIDIA technology inside them.

EVGA's range reached across motherboards, power supplies, cases, water blocks, closed-loop cooling, tuning software, keyboards, and mice. I remember branded memory, too, alongside laptops and PC-over-IP devices. Andrew wanted a consistent identity across those products. The SC17, designed in-house for overclocking, showed what that could look like as a complete machine.

That shared aesthetic and standard of service gave customers a reason to recognize and choose EVGA. Around NVIDIA's graphics hardware, the range offered a distinct brand presence that could extend across the rest of a computer.

That is the opportunity I still see: an in-house enthusiast brand that could carry NVIDIA's technology into complete systems with a consistent identity, strong customer support, and a loyal following. EVGA had already built much of what made that possible.

The people mattered as much as the catalog. Taipei had a deep pool of engineering expertise, and people moved between companies. Andrew had a gift for bringing that talent together. Their abilities could continue elsewhere; the team he assembled at EVGA had a value of its own.

Product context: EVGA's SC17 launch announcement, reproduced by TechPowerUp, the SC17 manual, and EVGA's product-manual archive. The catalog developed over time.