Who is amd ceo

She was promoted to senior vice president and general manager of Freescale's networking and multimedia group. In that role, she was responsible for global strategy, marketing, and engineering for the embedded communications and applications processor business. She left in to join AMD, also in Austin, as senior vice president, overseeing the company's global business units. Two years later she was appointed president and CEO, the first woman to run a Fortune semiconductor company.

But of course it's nowhere near enough and it's nowhere near fast enough. There's always much more that can be done. AMD's business is booming, and Su is credited with expanding the market for the company's chips beyond PCs to game consoles and embedded devices. AMD released products in with its Ryzen desktop processors and Epyc server processors for data centers.

They are based on its Zen microarchitecture, which enabled the chips to quickly process more instructions than the competition. The Radeon line of graphics cards for gaming consoles debuted in Today AMD is focused on building the next generation of supercomputers—which Su says will be "important in many aspects of research going forward.

Predicted to be the world's fastest when it goes into service in , El Capitan is expected to expand the use of artificial intelligence and machine learning. There currently is a tightness in the semiconductor supply chain, Su acknowledges, but she says she doesn't think the shortage will fundamentally change what the company does in terms of technology or product development.

And the products or technical decisions that we're making today will affect our products three to five years down the road. The semiconductor industry has never been more interesting, she says, even with Moore's Law slowing down.

Moore's Law, she says, "requires all of us to think differently about how we get to that next level of innovation. And it's not just about silicon innovation. It's also about packaging innovation, system software, and bringing together all those disciplines. There's a whole aspect to our work about just how to make our tools and our technologies easier to adopt.

The COVID pandemic has brought technology into the center of how people work, live, learn, and play, she notes.

Su was recently appointed to serve on the President's Council of Advisors on Science and Technology , a group of external advisers tasked with making science, technology, and innovation policy recommendations to the White House and President Biden. We all believed it was where people wanted to share their research.

It's just a great organization to move the industry forward. Kathy Pretz is editor in chief for The Institute, which covers all aspects of IEEE, its members, and the technology they're involved in. She has a bachelor's degree in applied communication from Rider University, in Lawrenceville, N. Please correct this typo: "Looking back, Lu Su, and I want you to spell her name correctly every single time! To fight on tomorrow's more complicated battlefields, militaries must adapt commercial technologies.

In August , engineers from Lockheed and the U. Army demonstrated a flying 5G network, with base stations installed on multicopters, at the U. Driverless military vehicles followed a human-driven truck at up to 50 kilometers per hour. Powerful processors on the multicopters shared the processing and communications chores needed to keep the vehicles in line.

It's , and the sun beats down on a vast desert coastline. A fighter jet takes off accompanied by four unpiloted aerial vehicles UAVs on a mission of reconnaissance and air support. A dozen special forces soldiers have moved into a town in hostile territory, to identify targets for an air strike on a weapons cache. Commanders need live visual evidence to correctly identify the targets for the strike and to minimize damage to surrounding buildings. The problem is that enemy jamming has blacked out the team's typical radio-frequency bands around the cache.

Conventional, civilian bands are a no-go because they'd give away the team's position. As the fighter jet and its automated wingmen cross into hostile territory, they are already sweeping the ground below with radio-frequency, infrared, and optical sensors to identify potential threats. On a helmet-mounted visor display, the pilot views icons on a map showing the movements of antiaircraft batteries and RF jammers, as well as the special forces and the locations of allied and enemy troops.

While all this is going on, the fighter jet's autonomous wingmen establish an ad hoc, high-bandwidth mesh communication network that cuts through the jamming by using unjammed frequencies, aggregating signals across different radio channels, and rapidly switching among different channels. Through a self-organizing network of communication nodes, the piloted fighter in the air connects to the special forces on the ground. As soon as the network is established, the soldiers begin transmitting real-time video of artillery rockets being transported into buildings.

The fighter jet acts as a base station, connecting the flying mesh network of the UAVs with a network of military and commercial satellites accessible to commanders all over the world. Processors distributed among the piloted and unpiloted aircraft churn through the data, and artificial-intelligence AI algorithms locate the targets and identify the weapons in the live video feed being viewed by the commanders.

Suddenly, the pilot sees a dot flashing on the far horizon through his helmet-mounted display. Instantly, two of the four teammates divert toward the location indicated by the flash. The helmet lights up a flight path toward the spot, and the pilot receives new orders scrolling across the display:. The two UAVs that have flown ahead start coordinating to identify the location of hostile forces in the vicinity of the downed aircraft.

A Navy rescue helicopter and medical support vessel are already en route. Meanwhile, with the fighter jet speeding away on a new mission, the two other UAVs supporting the special forces squad shift their network configuration to directly link to the satellite networks now serving the base-station role formerly played by the fighter jet. The live video feed goes on uninterrupted. The reconfigurations happen swiftly and without human intervention. Warfare has always been carried out at the boundary between chaos and order.

Strategists have long tried to suppress the chaos and impose order by means of intelligence, communication, and command and control. The most powerful weapon is useless without knowing where to aim it.

The most carefully constructed plan leads nowhere if it is based on bad intelligence. And the best intelligence is worthless if it arrives too late. The next key enabler is fifth-generation 5G wireless communications. These are sizable and complicated projects, and several different strategies are already becoming apparent. At Lockheed Martin , we're enhancing standard 5G technologies to connect the many platforms and networks that are fielded by the various branches of the armed services.

We call this our 5G. MIL initiative. Earlier this year, in two projects, called Hydra and HiveStar , we demonstrated the feasibility of key aspects of this initiative. Hydra yielded encouraging results on the interoperability challenge, and HiveStar showed that it was possible to quickly construct, in an area with no existing infrastructure, a highly mobile and yet capable 5G network, as would be required on a battlefield.

The new work takes an unusual approach. It is a collaboration with commercial industry in which technology is transferred from the civilian to the military sector, not the other way around. Radar, rocketry, and nuclear energy got their starts in military labs, and it took years, even generations, for these technologies to trickle into consumer products.

But today, for fundamental technologies such as computing and communications, the sheer scale of private-sector development is increasingly beyond the resources of even the largest national defense agencies.

To deploy networks that are sufficiently fast, adaptive, agile, and interoperable, warfighters now have little alternative but to exploit commercial developments. No wonder, then, that the U.

To understand the significance of such a shift, consider how the United States got to this juncture. In 18th-century conflicts, such as the Revolutionary War , the only battlefield sensors were human eyes and ears. Long-distance communication could take days and could be interrupted if the messengers it relied on were captured or killed.

Tactical battlefield decisions were signaled by flags or runners to commence maneuvers or attacks. By World War II, combatants had radar, aircraft, and radios to sense enemy planes and bombers up to 80 miles ahead. They could communicate from hundreds of miles away and prepare air defenses and direct fighter-interceptor squadrons within minutes.

Photoreconnaissance could supply invaluable intelligence—but in hours or days, not seconds. Today, the field of battle is intensively monitored. There are countless sensors on land, sea, air, space, and even in cyberspace.

Jet fighters, such as the F, can act as information-processing hubs in the sky to fuse all that data into a single integrated picture of the battlefield, then share that picture with war fighters and decision makers, who can thus execute command and control in near real time. Three Lockheed Martin military aircraft, built in different eras, have different communications systems designed to make it hard for an adversary to detect a transmission.

In a project called Hydra, engineers used electronic systems called open-system gateways to enable the three to communicate freely. From the top, the aircraft are the F, the U-2S, and the F Lockheed Martin. At least, that's the goal. The reality often falls short. The networks that knit together all these sensors are a patchwork.

Some of them run over civilian commercial infrastructure and others are military, and among the military ones, different requirements among the different branches and other factors have contributed to an assortment of high-performance but largely incompatible communication protocols.

Messages may not propagate across these networks quickly or at all. Here's why that's a problem. Say that an F detects an incoming ballistic missile. The aircraft can track the missile in real time. But today it may not be able to convey that tracking data all the way to antimissile batteries in time for them to shoot down the projectile.

That's the kind of capability the 5G. MIL initiative is aiming for. There are broader goals, too, because future battlefields will up the ante on complexity. Besides weapons, platforms, and gear, individual people will be outfitted with network-connected sensors monitoring their location, exposures to biochemical or radioactive hazards, and physical condition.

To connect all these elements will require global mesh networks of thousands of nodes, including satellites in space. The networks will have to accommodate hypersonic systems moving faster than five times the speed of sound, while also being capable of controlling or launching cyberattacks, electronic warfare and countermeasures, and directed-energy weapons.

Such technologies will fundamentally change the character and speed of war and will require an omnipresent communications backbone to manage capabilities across the entire battlefield. The sheer range of coordinated activities, the volume of assets, the complexity of their interactions, and their worldwide distribution would quickly overwhelm the computing and network capabilities we have today.

The time from observation to decision to action will be measured in milliseconds: When a maneuvering hypersonic platform moves more than 3. Our 5G. MIL vision has two complementary elements. One is exemplified by the opening scenario of this article: the quick, ad hoc establishment of secure, local networks based on 5G technology. The goal here is to let forces take sensor data from any platform in the theater and make it accessible to any shooter, no matter how the platform and the shooter each connect to the network.

Aircraft, ships, satellites, tanks, or even individual soldiers could connect their sensors to the secure 5G network via specially modified 5G base stations. They could also share data via military tactical links and communications systems. In either case, these battlefield connections would take the form of secure mesh networks. In this type of network, nodes have intelligence that enables them to connect to one another directly to self-organize and self-configure into a network, and then jointly manage the flow of data.

Inside the hybrid base station would be a series of systems called tactical gateways, which enable the base station to work with different military communication protocols. Such gateways already exist: They consist of hardware and software based on military-prescribed open-architecture standards that enable a platform, such as a fighter jet made by one contractor, to communicate with, say, a missile battery made by another supplier.

The second element of the 5G. MIL vision involves connecting these local mesh networks to the global Internet. Such a connection between a local network and the wider Internet is known as a backhaul.

In our case, the connection might be on the ground or in space, between civilian and military satellites. The resulting globe-spanning backhaul networks, composed of civilian infrastructure, military assets, or a mixture of both, would in effect create a software-defined virtual global defense network. The software-defined aspect is important because it would allow the networks to be reconfigured—automatically—on the fly. That's a huge challenge right now, but it's critical because it would provide the flexibility needed to deal with the exigencies of war.

At one moment, you might need an enormous video bandwidth in a certain area; in the next, you might need to convey a huge amount of targeting data. Alternatively, different streams of data might need different levels of encryption. Automatically reconfigurable software-defined networks would make all of this possible. The military advantage would be that software running on the network could use data sourced from anywhere in the world to pinpoint location, identify friends or foes, and to target hostile forces.

Any authorized user in the field with a smartphone could see on a Web browser, with data from this network, the entire battlefield, no matter where it was on the planet. We partnered recently with the U. Armed Services to demonstrate key aspects of this 5G. MIL vision. In March , Lockheed Martin's Project Hydra demonstrated bidirectional communication between the Lockheed F and F stealth fighters and a Lockheed U-2 reconnaissance plane in flight, and then down to ground artillery systems.

This latest experiment, part of a series that began in , is an example of connecting systems with communications protocols that are unique to their mission requirements.

All three planes are made by Lockheed Martin, but their different chronologies and battlefield roles resulted in different custom communications links that aren't readily compatible. Project Hydra enabled the platforms to communicate directly via an open-system gateway that translates data between native communications links and other weapons systems. Emerging technologies will fundamentally change the character and speed of war and will require an omnipresent communications backbone to manage capabilities across the entire battlefield.

It was a promising outcome, but reconnaissance and fighter aircraft represent only a tiny fraction of the nodes in a future battle space. But the X has been elusive at times. Indeed, a combination of new product launches, the pandemic, and cryptocurrency mining created a perfect storm for a lingering shortage.

The pandemic in particular "has just taken demand to a new level," Dr. Su said. More people found themselves working and educating from home, which drove demand for laptops and desktops, and the components that power them.

AMD doesn't actually manufacture any of its own chips. It designs them, and then taps outside fabs such as TSMC, which is investing billions of dollars into new and upgraded chip manufacturing facilities. And therein lies the reason for optimism. These investments were started perhaps a year ago," Dr. By Dr. Su's estimation, the chip shortage should begin to see some relief in the second half of Her prediction is generally in line with previous comments she made , when she indicated this past summer that AMD through its manufacturing partners was "bringing on more capacity every quarter.

The flip side to that is the shortage will continue through the rest of this year, which she previously indicated would remain "quite tight," and into the first half of next year, which she now predicts will be "likely tight.