Altera CEO Sees ‘Huge Opportunity’ To ‘Grab Share’ From AMD After Intel-Silver Lake Sale

‘Xilinx is being neglected, just like Altera was neglected inside Intel, so from that point of view, I see this [as a] huge opportunity to go and grab share, which is rightfully ours, and we are focused on doing that,’ Altera CEO Raghib Hussain says of his firm’s new independence.

With Intel’s majority stake sale of its Altera programmable chip business to private equity firm Silver Lake completed last Friday, the CEO of the freshly spun-off company said he sees “huge opportunity” to take FPGA market share from AMD.

Raghib Hussain, who became CEO of Intel spin-off Altera in May, told CRN in an interview a day before that he believes the company can ramp up competition against its larger rival because its newfound independence will allow it to move faster and put greater focus on FPGA chips, otherwise known as field-programmable gate arrays.

[Related: RealSense Almost Died At Intel. Now It’s Spinning Out With $50M In Funding.]

“We are going to move at a much faster pace, even faster than a standard silicon company out there, and being independent gives us that ability to do that, and we are going to be super focused on the customer, with extreme ownership—without any excuses,” said Hussain, who co-founded and sold chip design firm Cavium to semiconductor company Marvell Technology, where he served as a top executive for nearly seven years.

The CEO—who succeeded Sandra Rivera, the leader originally chosen to lead the Altera spin-off—said he believes Altera can thrive in contrast not only to the way the firm operated under Intel for several years but also to how AMD’s rival FPGA business has been run since the larger company acquired Xilinx for $49 billion in 2022.

“Xilinx is being neglected just like Altera was neglected inside Intel, so from that point of view, I see this [as a] huge opportunity to go and grab share, which is rightfully ours, and we are focused on doing that,” Hussain said.

Salil Raje, senior vice president and general manager of AMD’s Adaptive and Embedded Computing Group, said in a statement to CRN that AMD’s embedded business “isn’t a legacy FPGA business—it’s a strongly profitable engine for AMD.”

“It’s a portfolio that tightly integrates adaptive FPGAs with CPUs and GPUs into unified compute, not silos, which is exactly what our customers need and want,” he added. “Our roadmap is incredibly strong, and we’re winning major designs across critical markets. As AI continues to move beyond the cloud into the real world at the edge, that combination isn’t optional, it’s essential–and exactly what only AMD can deliver.”

Intel Monday said that it completed its 51 percent stake sale of Altera worth approximately $3.3 billion to Silver Lake last Friday, with Intel owning the remaining 49 percent. Altera’s revenue for the first half of 2025 was $816 million, up 16 percent from the same period last year, according to Intel. The FPGA firm’s gross margin was 55 percent while it had $356 million in operating expenses for the first six months of the year.

The Silver Lake deal—which valued the FPGA design business at nearly $8.8 billion when it was announced back in April—is the result of a plan that was supported by Intel CEO Lip-Bu Tan but originally hatched by his predecessor, Pat Gelsinger, back in 2023.

At the time, Gelsinger sought to spin off Altera, which Intel acquired for $16.7 billion in 2015, as part of his move for Intel to focus on its core business, namely its CPU design and contract chip manufacturing divisions. The original plan called for Intel to seek private investments for Altera, which was known as the Programmable Solutions Group, and do an initial public offering for the unit by 2026.

While Altera still plans to become a public company, it likely won’t happen next year, according to Hussain. Instead, he expects it to happen up to three years or so from now.

“While that is the goal, I'm not in a hurry because I think we can create more value for our customers and investors as well if we are a bit patient,” he said.

In the meantime, Hussain will continue with the new strategy Altera adopted when it became independent within Intel last year, which is to expand into the broader FPGA market beyond the focus it had on the telecommunications and data center segments under Intel’s ownership as well as to seize on the physical AI opportunity.

But rather than promoting FPGAs as an alternative to GPUs for AI needs, Hussain said the programmable chips make more sense as complementary products that are better equipped to convert and process data coming from physical sources like sensors.

“Right now, we are at an interesting time. The FPGA market is growing, and [it’s being] driven by this edge AI application where now you need to bring so many types of sensors and different types of data formats and I/O, [then] convert it into a unified form for a GPU to consume and so on. So the applicability of that is also increasing,” he said.

To enact his strategy, Hussain said he has made some big organizational and cultural changes, including bringing in his own leaders for sales and engineering. But he admitted that the fast-paced culture he is pushing may not be for everyone.

“I’m being very honest and transparent right from the first day that, ‘Hey, we are planning, as an example, to run the marathon in the Olympics [for] three to four years and win the gold medal. That’s what our goal is, and if you are on board for that, we would love you to be part of that.’ But at the same time, I respect that not everybody wants to run a marathon in the Olympics and try to win the gold medal,” he said.

On the sales side, he said the company is investing more into distribution and channel partnerships because he views them as key to serving the broader FPGA market.

“We are going to invest heavily in training their [field application engineers] and expanding our channels because those are the FAEs who will deal with the roughly 10,000 customers, which is [the] main FPGA market,” he said.

What follows is more of CRN’s interview with Hussain, who talked about why he thinks Altera has an advantage against AMD’s FPGA business, how its strategy is changing from how it operated under Intel, what the AI opportunity looks like for FPGAs, what the company’s timeline for going public looks like and how it’s investing in the channel.

How Altera Will Benefit From Its Newfound Independence

The world has become much more fast-paced. And I think that’s what we will bring with Altera: a much more fast pace as a part of [being] independent. That is the thing which most excites me about having this independence now because we are becoming [an] independent company. Silver Lake is taking 51 percent, which means full control and full independence to do things while keeping Intel as a great partner and customer because we do have product in [Intel Foundry Services], so I want to get preferential treatment, and having 49 percent [ownership] by Intel helps that. And of course, they are [a] large customer too, so that is also very helpful.

But being independent, it gives us full freedom on what market we want to focus on and how do we handle the customer, what product portfolio we want to invest in and which area we need to emphasize. So it really brings focus.

And the whole purpose of our existence is our customer. That’s how I have done [things] throughout my life. As you know, I was co-founder of [chip designer] Cavium, which we built [into a] reasonably sized company, we sold it to Marvell [Technology], and that’s how I became president of [products and technologies] at Marvell. And when [I was at] Marvell, for seven years, I led data center infrastructure, basically. So through organic [and] inorganic investment, we were able to drive huge growth for data center. So basically, when I entered Marvell, it was $20 million revenue in data center, and this year is going to be north of $5 billion, so it’s huge growth for that. And that all comes from the focus and execution and doing what the customer wants to do so that you solve their problem, you make them happy, and that’s how it’s a win-win situation.

So that’s the same culture I’m trying to drive here: a culture of ownership, extreme ownership, a culture of focus, a culture of value creation, a culture of working together, knowing that we are all together for success and failure. There are no independent silo rows and all that. So those are all the things I’ve been doing for the last four months here, and with independence, I’ll have a lot more freedom to drive it more aggressively.

How Altera’s Strategy Is Changing From Its Days Under Intel

First of all, I joined Intel before [Intel’s majority stake sale of Altera to Silver Lake was completed]. Normally, CEOs join after [the deal closes], right? But I thought, if I go earlier, I will be able to start making changes the way we want to do it.

So from the product and market point of view, under Intel, for all the right reasons, they were more focused on telco and data center, right? Because it’s more of an x86 [CPU] offload. Think of it that way. And that is why the product portfolio and customer focus and sales focus was not on the broader FPGA market.

The FPGA market is a very established market, and that is very broad also, ranging from test and measurement to audio-visual to aerospace-defense to industrial embedded. So many applications because those are the applications where FPGA is the best fit. FPGA basically is good when protocols are changing and all that, and you want flexibility but at the same time hardware performance. And then also when the volume of the application is not such that you can justify doing an ASIC [application-specific integrated circuit]. Because the cost of ASICs is growing, the applicability of FPGAs is growing.

Right now we are at an interesting time. The FPGA market is growing, and [it’s being] driven by this edge AI application, where now you need to bring so many types of sensors and different types of data formats and I/O, [then] convert it into a unified form for a GPU to consume and so on. So the applicability of that is also increasing. So it’s an interesting time where the broader FPGA market—[which is being] driven by AI—it is growing. So for us, it is very important to focus on that and drive that. So that is what we are going to do.

Hussain On How His Focus On Culture Will Help It Compete

As I said, I have always developed and done business circling around and focused on the customer, so that’s the culture I am driving. That’s a mindset I am driving: extreme ownership. And that’s why some of the leaders that I am bringing [have] the same mindset, same style. So you must have noticed I have brought a new head of sales from Xilinx. I brought a new head of engineering, who was at Xilinx and left Xilinx, went to Marvell and is now with us. I am also in the process of bringing [in the] rest of the executive staff.

So this is the team that I am building, and I’m being very honest and transparent right from the first day that, ‘Hey, we are planning, as an example, to run [the] marathon in the Olympics [for] three to four years and win the gold medal. That’s what our goal is, and if you are on board for that, we would love you to be part of that.’

But I do understand, at the same time, that not everybody wants to run the marathon in the Olympics and try to win the gold medal, right? So that’s fine, and people can self-select them out. And as a result of that, of course, a lot of people—we had a lot of changes, as you may have noticed recently on my executive staff as well as in the broader company as well. So I expect that. That is fine. It’s a different culture. It’s a different goal. We are on a mission here, and not everybody in life wants to be on a mission. So that’s perfectly all right.

Why Hussain Thinks Altera Has An Advantage Against AMD

We have a core talent team, core product. We have a leading fabric in the industry right now. We have a leading compiler latency time. For example, we have improved a lot on the overall software. There are only two companies who have such a good software suite for the FPGA, which is [AMD’s FPGA business, formerly known as] Xilinx and us. And most importantly, we have this talented team who is actually super excited right now to be able to do again what they want to do. And then the other folks who are also FPGA-focused, people who want to do FPGA and [are] now at AMD Xilinx, they are [finding themselves to be], a little bit, second-class citizens. They are actually jumping ship. A lot of them are joining us. So it’s a very exciting time building this new focus [and] new team.

And, in reality, Xilinx is going through similar issues that Altera went through under Intel—and for the right reasons, by the way. If I was AMD’s CEO, I would think the same way because no matter how much Lisa [Su] invests in Xilinx, she cannot double [its] revenue because they already have 70 percent market [share]. So how much can you grow? An additional billion dollars in revenue is not that useful in a bigger picture.

So if I am CEO of AMD, I’ll [take] my best engineer, pull them from the FPGA side and put them on the AI side. Similarly, I’ll pull all my good sales folks, the support folks from the Xilinx side and put them on the AI side because this is where the biggest bang for the buck is, right? So nothing against Xilinx in its management by AMD. But if they focus on Xilinx and not on AI, that is going to be disaster, right? So they are focused on what is right for them. But as a side effect, of course, Xilinx is being neglected, just like Altera was neglected inside Intel. So from that point of view, I see this [as a] a huge opportunity to go and grab share, which is rightfully ours, and we are focused on doing that.

How Hussain Views The AI Opportunity For FPGAs

In AI, two very specific things are happening. One is, when AI meets the physical world, whether it’s robots or whether it’s medical equipment using AI, doing certain special things and so on. There are many, many sensors now: the pressure sensor, the temperature sensor, the cameras and LIDAR and radar. All those have their own special interfaces, and they all have a special data format, and FPGAs traditionally have been [the] best [way] to convert all those things, being a kind of glue logic, bringing all [that] data, putting it in a unified format and presenting it to the GPU or ASIC and so on.

Somebody can argue why GPUs cannot do data transformation. Yes, they can. It’s just too costly. Because the fact is, would you rather use GPU for GPU work, or would you use GPU for doing basic data transformation? [In the] same way, somebody can argue, why can ASIC not do it? Yes, ASICs can do it. But the problem is this: If you keep trying to implement a specialized ASIC for each kind of these applications, for a specific type of OS, you’ll have to implement many, many ASICs. So the easier path is this: You implement ASIC for the purpose that you want to implement ASIC, and [it] will be complemented with FPGA, which will translate different types of [input and output data for] the ASIC based on the application.

So [using] AI with the physical world, the applications are going to be very, very specialized. For example, the type of interfaces and sensors that you need in humanoid robots is completely different compared to the type of thing that you need in drones, compared to the type of thing you need in agricultural AI robots, compared to the thing that you need in medical equipment, which is doing your [gastrointestinal] scan and processing the AI right there.

So there are so many things. A lot of people are saying, ‘Oh, we are saying that FPGA will [replace] the GPU.’ No, that’s not what we are saying. What we are saying is this: FPGA is the best [complementary] solution as a glue logic for the ASIC and for the GPU.

Hussain’s Timeline For Altera To Go Public

The goal is to take the company public at some point. However, I’m not in a hurry. What I mean by that is being private gives us a lot of flexibility to invest in the markets and the product that we want to invest [in] for a healthy business. Once you become public, you have a lot of constraints in what you can invest in and not in how much can you invest.

So my goal is to make the company and business very healthy first [by] having a robust portfolio and having the design wins and the [customer] pipeline and all that established. I think it will take a year or two to get there, for sure.

The good part of this [is] my partner is Silver Lake, who believes in long-term investment. So they are not in a hurry to cash out or any such thing. So they are here at least for five years or more. If you look at [their] track record in Broadcom, they have been [invested] for 20-plus years. In Dell, they have been [invested] 10-plus years. Even in NXP [Semiconductors], they stayed for a long time, 10-plus years. So they have that understanding. They do understand the FPGA market takes time, so the product that we are building today will go and generate the volume revenue in five to seven years. They do understand that.

So I think while that is the goal, I’m not in a hurry because I think we can create more value for our customers and investors as well if we are a bit patient and all that. But I think it will be around three years [for a] time frame or so.

How Altera Is Investing In Channel Partners And Distribution

This is one of the areas where, under Intel, the focus was moved from away from the channel and distribution partner. So the application of the FPGA is in the broader market, and the broader market you can only reach through that distribution and channel. So we have an absolute focus on that because that is an area we need to grow [the] business mostly.

The direct major customers, I have a very good relationship with all data center and all telco [customers]. I really do not need [the] sales team to engage over there because I believe in engaging directly with the customer.

The focus of the sales team that we are building, a new sales team here, is distribution and channel, and I myself am focused on that. We are going to invest heavily in training their FAEs [field application engineers] and expanding our channels because those are the FAEs who will deal with the roughly 10,000 customers, which is the main FPGA market.

How Altera’s Business Has Been Trending Recently

The broader market is [seeing] kind of incremental growth. There’s not that much growth because of all the inventory issues and so on and so forth, [for] which we have reached the bottom, but it’s still broader. If you look at the semiconductor industry reports, most of the folks who do not have exposure into [the] data center [market] are not really growing. But the major growth, which we have right now, is driven by the data center. We have exposure in various data centers. We have solutions [for those], and because they’re driven by AI, those are ramping very well. Last year Altera’s revenue under Intel was $1.5 billion. Of course, we have not publicly shared our number for this year, but let me put it this way: It is reasonably higher than last year. [Intel said on Monday that Altera’s revenue for the first half of 2025 was $816 million, up 16 percent from the same period last year.]