Back in January Moshe Gavrielov announced that he was leaving Xilinx. He was replaced as president and CEO by Victor Peng, who had been the company’s chief operating officer.
“Gavrielov’s decision to step down brought to end a remarkable career, spanning 40 years in semiconductor and software related companies,” said Peng.
Peng’s appointment, after ten years with the company, comes as Xilinx, which invented the FPGA, hardware programmable SoCs and now the ACAP platform, looks to embrace a new vision and strategy. According to Peng, it will bring “new growth, technology and direction,” to the company.
Crucially, Peng sees the company moving beyond the traditional world of the FPGA to one that addresses and enables a “more adaptable and intelligent world,” where Xilinx will deliver, “a new category of highly flexible and adaptive processors and platforms.
“We want to be able to support rapid innovation across a wider array of technologies,” Peng explained,” from the end-point to the edge.”
According to Peng three trends are affecting the market at present.
“The first is the explosion in data which is driving demand for a different type of computing; the second is the advent of artificial intelligence which will disrupt multiple industries.
“The third is exponential change. There’s currently no single architecture capable of addressing the complexity that comes with that. In truth the economics of Moore’s Law no longer work, and they haven’t for some time.”
But, despite these trends, Peng believes that for those looking for them, plenty of opportunities exist.
“Things are changing so dynamically that you need a vision and strategy to manage this intelligent connected world.”
How you formulate and then deliver that strategy is crucial.
“Everything we do has to be focused on delivering a strategic plan,” explains Peng. “The transformation of any business isn’t going to happen overnight but we have been addressing these fundamental challenges for some time and I’m lucky to be building on the strong foundations that were left by my predecessor.
“There is a degree of continuation, in terms of our core business where we have some real strengths, but going forward we see the biggest opportunities coming from developments in data centres.”
According to Peng, ”When we talk about strategy, I’m talking about a five year time line in which we will be focusing on delivering our plans for the business.”
Strategic plan
Peng describes the company’s strategy as comprising of three core elements.
“The first, and most important is, as previously mentioned, a new emphasis on data centre acceleration,” Peng explains.
“At Xilinx we are looking to ramp up our efforts with key data centre customers, ecosystem partners and software application developers. We want to enable innovation and deployments in compute acceleration, computational storage and network acceleration.”
Peng points to the data centre market as one in which there is rapid growth. Customers are looking to adopt new technologies quickly as they look to take advantage of the orders of magnitude performance and performance per-watt improvements that will be required to enable applications like artificial intelligence (AI) inference, video and image processing, and genomics to be taken up.
“While we look to transform the business, there will be a degree of continuity going forward,” Peng suggests. ”Xilinx has many thousands of loyal customers and while we are looking to exploit new market opportunities, and I believe that we have reached a significant inflection point, we can’t ignore their needs either. These core markets and customers remain central to Xilinx and we will continue to enable innovation in these areas.”
That is the second essential pillar of the company’s strategy going forward - accelerating growth in its existing core markets.
“These markets are areas where we have been a key technology leader and have developed deep market traction,” Peng explains.
“Traditionally we have served multiple markets, primarily wireless communications and infrastructure but also audio, video and broadcast; aerospace and defence: industrial, scientific and medical and consumer technologies.”
In time, according to Peng, “We want to make the data centre market our number one segment and much of our focus will be on delivering that ambition.”
Xilinx isn’t alone in targeting the data centre market. Intel PSG, for example, now describes itself as a datacentric business and talks of broadening its portfolio of programmable logic solutions across its silicon, software, intellectual property and hardware platforms.
“If Intel has ‘pivoted’ to become a datacentric business so too has Xilinx,” argues Peng. “The world is changing and data is taking centre stage.”
Peng, however, believes that Xilinx is better positioned to address these market changes than most of its competitors.
“We are in the end point, whether that’s in drones, cameras or vehicles and we have developed a broad range of cost optimised, low power devices such as Spartan or our Zynq SoCs. Not only that we are well placed in communications infrastructure and to address the acceleration that we are seeing into the Cloud.”
The third element of the company’s strategy has been the introduction of the Adaptive Compute Acceleration Platform (ACAP).
“I would argue that this is the most significant pillar in terms of our strategy and this new product can honestly be described as a ‘breakthrough’,” Peng contends.
ACAP is a highly integrated multi-core heterogeneous compute platform that can be changed at the hardware level to adapt to the needs of a wide range of applications and workloads.
“That adaptability, which can be done dynamically in milliseconds during operation, delivers levels of performance and performance per-watt that is unmatched by CPUs or GPUs, and is crucial,” Peng says and it means that, “While FPGA and Zynq SoC technologies are still core to our business, Xilinx is no longer just an FPGA company.”
The company’s strategy appears to be paying dividends.
Announced last month, the company has been able to post record revenues of $2.54 billion for the fiscal year 2018, up 8% from this time last year.
“Our focus on data centres; meeting the demands of new and emerging applications and the introduction of the ACAP platform seems to suggest that our three-pronged approach is starting to work,” Peng concludes.
