Arm designs instruction sets for modern chips, and Qualcomm, Apple, and Samsung chips are all Arm chips. Arm licensed the instruction set to these companies, which then began producing chips with various custom capabilities. This model has been a great success.
Renejas, Arm's new CEO, spoke to the verge and answered some questions: How does Arm make money? Who are its customers? Even if it doesn't actually make chips, is it affected by chip shortages?
Arm is a key piece of the puzzle across the tech ecosystem. We all rely on Arm. It affects all of us, but it can be very opaque. Consumers don't really have a direct relationship with Arm, it's mediated by many other companies. So the first question is, what exactly does Arm do?
Arm is not a well-known company and is not well known to the public, but we think we are very important. You can find Arm technology in virtually any type of semiconductor and/or OEM product. We are in smartphones, laptops, smart TVs. Looking around my desk, there are probably dozens of Arm processors everywhere. We're not actually building anything. We design for all parts of the product, which is intellectual property. We build that design, not build chips, and we license that design to the people who will build the final product. Our most famous main product is the CPU, which is the brain of almost all types of electronic devices.
We don't make chips, we empower our brains to other people who are willing to make chips. Based on that, there are a lot of Arm brains out there. Looking back at the last quarter's data, 7.4 billion chips were manufactured and delivered across all semiconductor companies and OEMs worldwide, including some type of CPU, GPU or technology from Arm. That's a huge number.
Who are Arm's customers?
In the electronics industry, it's easier to say who isn't Arm's customers. Almost any company you can think of is our customer. We have TSMC, Samsung, GF — these are all companies that actually make chips. Intel, AMD, Nvidia, Qualcomm, Amazon, Microsoft and Google are also customers. For customers in other parts of the world, we have Alibaba, Tencent and ByteDance.
Almost every tech company is our customer.
How do you make money from these customers? Are they just licensing designs? Are they buying off-the-shelf reference chips? Do they pay patent licensing fees? How does this work?
Our business model has two components. We have an upfront license fee that our partners pay us to get the right to use the technology. This gives them the right to design using our technology. If these designs end up in production and final product, we will charge a per-unit royalty according to some rules related to the contract. So, at a high level, there are two sources of income. One is what we call licensing revenue. The other is what we call royalty income.
If I go to Qualcomm and buy a Snapdragon chip, do I need to pay you? Or does Qualcomm have to pay you? Or did someone else give you money?
Personally not, but Qualcomm has. In this example, Qualcomm reports to us the shipped quantity purchased by the customer. There is a pre-negotiated royalty rate and they will pay us to pay these royalties.
Samsung makes and designs its own chips for its smartphones, and these divisions of Samsung actually have to sign up with each other. If I buy a Samsung phone with an Exynos chip — instead of a Qualcomm chip — how does that money flow back to you?
That's the beauty of Arm's business model. Qualcomm uses our technology, and so does Samsung. When the Galaxy phone ships, we'll most likely receive payment. If the product works with Qualcomm, we will get paid by Qualcomm, and in the case of Samsung, we are still getting paid. It depends on our licensing arrangements with our partners, but typically, they work with the semiconductor divisions of these companies. Taking Samsung as an example, if we sign a contract with the chip department, then the chip department will pay Arm.
Apple has obtained a very unique license called the Architecture License. They completely designed their own chips, and it seems that Apple does not use many Arm designs, but it is the intellectual property of Arm. How does Arm get paid when a consumer buys an M2 MacBook Air? Can anyone get the kind of license that Apple has, they can use your IP, but can they completely design their own things?
Commercially, it's very similar. We have contracts with companies like Apple that pay us royalties like everyone else.
The architecture license gives the Company the right to manufacture Arm-compatible processors. Anyone who gets permission can make minor modifications to the microarchitecture, which actually refers to how it is physically placed on the chip. But what they can't do is modify their CPU so that they can't run Arm instructions. This is very important because, at the end of the day, we have to maintain software compatibility. If someone is running an Arm processor, whether it's something we're building or something built by a partner with an architecture license, it must meet the requirements to run Arm software.
We don't have a lot of architecture licenses because it's hard to do. We make very good CPUs. It's very difficult to build a CPU that's different or better than what we've done, while still meeting the Arm standard. Not many people do this. There used to be a lot, but now there are fewer of them. These are teams that are hard to build. Most companies look at it and say: If I'm building an SoC and I only have so many valuable engineers to differentiate my product, a differentiated Arm CPU might not be the best option to spend time. Better places to go around areas that Arm doesn't do, like cameras, modems, or IOs.
When Apple or another architecture licensee delivers a product, does Arm have a team to verify that it is running the Arm instruction set and that they are not violating the rules?
We have a set of requirements and a compliance suite that we test to essentially verify that what they build meets Arm standards. We test it to see if it can run Arm directives and code, if the compiler is corrupted, or if you don't recognize the correct instructions. The short answer is that we do provide a set of compliance tests for anyone building a design based on architectural licensing.
I want to ask these questions to close the loop of this company in almost all modern chip centers. I think the closed loop is where the real validation is that this is the Arm way to make the Arm architecture valuable.
Both Arm implementations and architecture implementations run software for the Arm Instruction Set Architecture (ISA) without interruption, allowing customers to add custom instructions. While this sounds innovative and cool, what really makes CPU architecture relevant over the long term is letting developers know that it will work. If a developer is writing a piece of code for an OEM and has an Arm embedded, the developer doesn't want to know, doesn't need to know, or even can't even know that they're designing a thermostat with a Bob chip inside, and Bob has some extra instructions. You need to take advantage of this, as they may not know if other OEM devices have an Arm chip that contains these instructions. It's important to level the playing field and ensure that software data sets look the same. Our founders have done a fantastic job of sticking to this and making it have the principle of speed. You can see that it really benefits us now.
Arm has a lot of customers, and they all compete with each other in various ways. A lot of executives in positions like you feel like they do a lot more political work than engineering or research. How do you tell the difference? Do you think you're a politician and just putting everyone on a level playing field? Or are you someone who goes deep into processor design?
We're definitely bogged down in processor design. I mean, that's what we do. In the end, we spent a lot of time and effort developing these CPU and software ecosystems, solving partner problems, and really making sure the product was ahead of the curve. One of the things that is a little different about us is the fact that we deal with everyone. We have to keep consistency in the way we manage our partnerships, which really revolves around access to technology, access to mistakes, and access to people. The world depends on Arm. As you said, we're a bit opaque. I can say on a podcast that the world depends on Arm, and someone listens to the response, "I depend on you?" I don't even know you! " We manage our relationships with our partners very carefully,
Arm is not an industry standard body, but a company that reports earnings, and if it wants to go public, it must increase earnings for shareholders. How did you handle it?
Arm tries to be neutral. Arm is known as the Switzerland of the electronics industry, and that's not a bad balance. We're not trying to pick winners. Arm participates in the ecosystem of the ecosystem. If you start at the very bottom of the semiconductor chain — GF, Samsung, TSMC, Intel, and all the people who make chips — you have to work with everyone. Arm had to ensure that our technology could be built on every semiconductor process in the world, which required the investment of all these partners. And then, all the way up, when you think about Android, Linux, Windows, and all the major operating systems that we support, we have to make sure we're there too.
Arm really tries not to take advantage between one person and another. I'm not sure how we'll end up doing that, but it means we're standing on the shoulders of a sizable ecosystem. I call it an ecosystem because there are design tools, manufacturing processes, software operating systems, and middleware. We often work with interest groups and create them. We didn't really work much with standards bodies, but we did work a lot with all the industry players I mentioned. Arm makes sure to understand everything they are trying to do from a roadmap perspective to ensure we are as compliant as possible. At the end of the day, we license the technology to the people who build the chips. People who make chips want to have the widest range of opportunities where they make chips,
How many people work at Arm? What is their structure? Is it all chip design, or is there a lawyer? Is the ratio of patent attorneys to chip designers 1 to 1? How does this work?
The number of Arm people may be around 5,800, but it may be more than 6,000 people including outsourced people.
We are mainly engineers, most of them in the UK. Arm is headquartered in Cambridge. There are several different engineering sites in different parts of the UK, France and Northern Europe. There are several design centers in the United States – in Arizona and Texas. There are also a considerable number of design engineers in India: Bangalore and Noida. Most of Arm's employees are engineers. The legal department is very small.
Arm has done a great job of understanding how our licensing model works and how to protect its intellectual property. Arm doesn't have a huge legal department, but we do have a lot of engineers because these products are really hard to make.
Arm is known to be embroiled in a storm for the SoftBank Vision Fund, which raised billions of dollars. They invested in a lot of companies and bought Arm directly. SoftBank then tried to sell the Vision Fund to Nvidia when it was unstable. The industry basically lobbied against the deal to the best of its ability, saying, "We don't want Nvidia to have this core CPU technology processor, this design technology." The government said they would block the deal, and SoftBank gave up.
You arrive as the new CEO. You say, "We're going to get the company public." It was a big twist. Your ex was adamantly opposed to letting the company go public because the pressure to increase revenue after going public put at risk the model of being a neutral and fair supplier for everyone, as companies could increase revenue through special deals. These pressures will come, but you still decide to go public, how did you make that decision?
We announced this change at the end of last year when the Nvidia deal was essentially bankrupt. After I finished my fiscal year, we were finally able to talk about our financial performance. During the Nvidia era, we remained silent. When we announced revenue for the year, we set a record of more than $2 billion. Arm has a turnover of $2.6 billion and an operating margin of nearly 40 percent, but because we remain silent, the outside world believes that Arm is losing money. In fact, the business for the quarter was higher than we were at the end of last year: more than 50% and $700 million in revenue, of which $450 million was royalties.
Shortly after SoftBank acquired Arm, Arm restructured and created two business units. Arm began to turn to other markets. This fulcrum is not only a business model, but also a product. Arm knows something is happening in the data center. Arm has made good progress on software workloads. If we push our investment towards specific directives, such as SMEs and SVEs — these are vector extensions for specific workloads on hyperscale computers — we can gain something in the hyperscale computer market (i.e., the cloud computing market).
AWS is an important partner for us. They announced Graviton2 and some pretty jaw-dropping numbers that offer a 40% improvement in price-to-price performance over other architectures. Therefore, we diversify our business by solving this problem not only by developing different products, but also by different parts of the business model strategy. We know our business will be fine. All the financial results that you're seeing now, are fantastic, and the team has done a really good job of that, really from the work of a few years ago. You won't see royalty results for your unit overnight. We develop the IP, and the IP must be handed over to the customer. They have to make a chip, and that chip has to go into the product. Then the product must be qualified. All of this can take three to four years. We felt so good about where we were going. In the areas that Arm has been investing in, such as the cloud, automotive and the Internet of Things, we also feel very good. These are the big long-term growth areas that I think we're a good fit for.
Do you think inter-quarterly investor pressure will change the way you run your company? Because it seems to be a risk of going public.
I think Arm is a public company at any time, it's just part of the way the world works. I can't talk too much about what life as a public company would look like, though. I can say that we are very confident in the long-term growth of the markets involved.
You've just expressed your confidence in the automotive, the Internet of Things, and the cloud. The cloud market exists and grows. Arm has already made this move, and it seems certain to happen. As CEO of Arm, can you drive this? Or is it more like, "We're just going to design and product ready for the car to truly become a networked computer on wheels"?
When people think of electronics in cars, they automatically turn to autonomous driving and so on. That's just one dimension. Think about what's inside the car and the number of processes in the car. The dashboard, all digital, fully computerized, runs on almost everything on the Arm. The drivetrain, anything to do with mirrors and brakes, all of this is moving quickly to the Arm. Much of what happens in the car is moving toward the Arm, regardless of the dashboard or autonomous aspect.
In the power train, there are many older electronic control units (ECUs), which are separate units. They may have an old proprietary microcontroller, and they don't communicate with other parts of the car. All of these ECUs are being redesigned. Inside the car, there may be 50-70 ECUs. They have many, and each one may have an old proprietary microcontroller with no connections, no memory management unit, and no communication with the rest of the car. As a result, cars become highly connected devices. Then add automatic and advanced driver assistance systems (ADAS), an area where Arm is evolving.
Cars are a bit like a sandbox of multiple technologies: dashboards, drivetrains, drivetrains/ECUs, and autonomous vehicles. Going back to the data center, when it comes to computing, what's really important in the car is performance and efficiency. You can't have a server in your trunk running an electric car and succeed. Some of today's cars, that's it. They're a bit like servers in the trunk. Over time, things will get better, so we're very optimistic about the automotive market. We are growing very fast there.
So would you ask some of your engineers to say, "Go figure out an ECU, engine controller, or body control module that works across the car so that someone at Nvidia can come and license it and sell it to Ford"?
This is already happening. What do these CPUs for cars matter? Efficiency, power and functional safety are all important. You must have all the redundancy to ensure that it can operate in a safe manner. Some people do this in software through a computing library, but most people prefer to do it in hardware because it is more secure and efficient. We have developed automotive processors and graphics processors that embed functional safety. We have not done so in the past. We're basically going to roll out a generic thing to use anywhere. This is one of the things that has changed significantly in the last few years. Cars will be a very big market for us.
Back in the present, we've been in a chip shortage for a long time, and that may or may not end. Intel and Nvidia's quarterly results were very bad. We don't make anything of the view, but we are very relevant to it because our royalties have to do with how many components people ship.
Going back to the previous discussion, we are very diverse in terms of end markets. Another thing that's happening is that more and more CPUs are being used in these SoCs. The application processor for a mobile phone may use one CPU, whereas the cluster used for computing is now 9 or 10 different CPUs. You have 10 CPUs, that's just the application processor. Then consider touch sensors or anything on a display or camera. We see that while sales in certain markets, such as smartphones, have been weakening, we have been protected by it from a product direction perspective.
I do think this one is a little different. It's not all technology, it's not all territory. The Internet of Things is strong, the industry is strong, the cloud is strong, and the different nodes are strong. Sometimes, older technologies — 14, 28, 40 nm — are a bit hard to come by. All the equipment in these complex systems needs to be mixed and matched. You only need one thing to be fine. Since COVID-19 and the world are not the same as they were three or four years ago, a supply chain doesn't take much time to screw up. Next you know, we're missing products like toilet paper. Some people will say, "How come we still lack toilet paper?" "These tissue mills have a very fixed set of equipment that makes commercial toilet paper for industry and enterprises or does residential toilet paper. When we all go home and no longer go to offices, malls and movie theaters, the demand for commercial toilet paper is out of balance with the needs of consumers. That's why the toilet paper thing happens.
So now, think of a car that needs all kinds of diodes, capacitors, resistors, and thermal sensors. This is not just because it is difficult to obtain 3 nm, but also because it is difficult to build a fab. In the past downturn in business cycles, people stopped developing and slowed down projects. They don't do new things, they stop innovating. I didn't see it this time. I think it's because everything is digital and there's a lot of innovation going on. What are our indicators? Licensing and new design starts have never been so good for us. It is available in all markets.
With the prospect of globalization being widely reconsidered around the globe, the world of semiconductors is less calm. The United States has just passed the CHIPS Act to really incentivize chip manufacturing and various design investments. TSMC is being built in the United States. Intel has just broken ground on a new plant in Ohio. Do you see that taking into account, "We need to transfer these critical dependencies to our country." Or are you just watching things?
We will certainly be involved in the dialogue. As long as we can expand the need for this by talking to political officials in any country, we will help. It's not just an Arm issue. This is an industry issue. A single point of failure for anything you're doing is not a good thing. 50 years from now, every continent should have a world-class fab.
One of the realities of how the world works today is that cutting-edge process nodes are largely controlled by TSMC. That's 5nm, 3nm. They are far ahead of it and control it. That's what they are now. Very few people can compete at this level. Older process nodes (40 nm, 14 nm) for automobiles are severely limited and no longer have anyone building these fabs. This is not a good investment situation, how do you see the results?
What I can tell you is that we're seeing a lot of work being done on how to convert certain fabs into new jobs. Like you said. People are often a bit reluctant to put a lot of money into the final minus of three fabs. You start thinking, "Can I convert the fab we're building into some kind of process technology and turn it into a logic fab?" Is there a memory factory that can be converted into a logic factory? But it's a very complicated issue because there are only so many factories in the world, and there are only so many people who know how to build these things. These companies are publicly traded. They have to make money. This is a very complex matrix.
Thirty years ago, Japan was world-class, with logic factories all over the country. This has basically dropped to almost zero. Japan is looking for ways to retrofit it to bring more fabs back online. I think you'll see more CHIPS type work done in the U.S. in other countries. South Korea is mostly a fab, but I think it also has huge capabilities. However, this is a complex and terrible issue.
How dependent is Arm on TSMC? It feels like a lot of companies realize that they really depend on a company. It's an industry leader for a reason. Have you considered this dependency when designing a new processor type or architecture?
We work closely with TSMC. I would answer that all fabs are important to us, back to my comment in Switzerland. In an ideal world, there are many people who can build all the cutting-edge technologies. In general, in smartphones, in data centers, this is often the cutting-edge process because people are really trying to get the best performance possible – not in automotive and industrial. Any fab with cutting-edge processes is very important to us.
As we all know, Intel has long been an integrated designer and fab company. They are basically surpassed by TSMC in the world. They now have a new CEO, and they're breaking the bandwagon. They said, "Okay, we're going to open our factory to other people." You just said Intel is the customer. Is this something you're working on with them?
We wanted TSMC to be able to make a lot of Arm products, and they did. We hope Intel will do the same. I think Pat Gelsinger did a lot of the right things. His job is hard, but at the same time, I think he's done some really good work and we're more than willing to do more with him.
On the chip bill, the bill was passed and signed, and everyone was excited. Intel immediately received a lot of criticism as it announced cuts to capital expenditures and increased dividends instead of putting them into fabs. Broadly speaking, do you think this is the right move? What kind of timeline should we look for for fabs in the United States?
I wouldn't doubt Pat Gelsinger's decision about how he spent his money. I think we need to move fast and get these facilities up and running as soon as possible. We need to get them online as soon as possible to reduce the risk of single points of failure. I think we need more fabs.
What timeline do you make your decision based on? It looks like if a bill were passed today, or if someone announced a new fab today, we wouldn't see it in five years. How do you see your decision-making schedule?
After six months as CEO, I'm still learning about CEO responsibility. One of my biggest responsibilities is to think about what will happen five years from now. In the way our business model operates, the time it takes us to design a product until we see revenue is quite long. We are thinking about where the opportunities are, where investments need to be made, where the threats are, and the direction relative to system and software design. We've been thinking about these things, we've been talking about them. Most of my energy is spent thinking about what the world will look like in 2025 and 2026, not the next quarter.
There are some long-standing bets that focus more on consumers than replacing servers in the trunk of a car. The hottest trend currently related to the Arm takeover is Apple's shift from Intel to Arm-based processors, the M-series processors in their own Macs, which is a huge win for them. Is this something Arm can do for other industries? Or, "This is the design. Hopefully, Qualcomm can make chips that Microsoft and Dell can do the same?" From your perspective, how do you see your relationship with the industry?
I think we can do more and probably need to do more. It's not because we need to help the industry, but because the builds of these products are really complicated. Let's start with architecture licensing. At some point in time, it was thought that architectural permission was needed to build a better CPU to compete with the Arm. The microprocessor's per-clock IPC instructions are just micro-fragments that can really tilt the design. You consider the memory subsystem, the interconnect, the size of the cache, how you interface with the rest of the SoC, and make sure that the SoC is in a multichip package with a bare to die interconnect.
Some customers are very good at these things and can figure it out, so they don't need a lot of help from us. But over time, more and more industries will benefit from Arm helping them do more. This is an area that we are following very closely.
There are opportunities in the future, and it is clear that the need to do the system on the chip will only become more and more difficult.
From your perspective, as long as they're all Arm customers, it doesn't matter who wins, right? If Apple has 100% of the laptop market, that's great for you. If Microsoft, HP, and Qualcomm can find competitive Windows on Arm laptops and they have a 50% share, you can still get paid. Does it feel like whoever wins, you win?
The way to think about it is that if the product ships with the Arm ISA, then this is a good thing for us. If an alternative device is used, it means that it does not use the ARM ISA, so this is not good for us. When I took on my last role, I gave me a nickname on the website and we worked with the PR team to basically say "Wherever the calculation happens, the Arm will appear". You can take a step back and say, "Oh, this is my thermostat or microwave, Polycom, smart TV, laptop" – whatever it is.
For most people, computing happens on their laptops and desktops. Windows on Arm has been around for 13 years, but it hasn't happened. Are you just content to let Apple take over the industry? Are you saying, "We have to push this," because there are other people's instruction sets — that is, Intel's — that dominate in this incredibly large industry?
Before joining Arm, I worked at Nvidia, which was my product line. We have made great progress in that regard. We don't have ports and Apple doesn't help us. Fast forward to 10 years before streaming audio has emerged. No one really cares about these apps. If you think about all the apps that run on your PC or phone, it's hard to say, "Okay, which native apps aren't really running on Arm?" Almost all of them were there. Stay tuned for your PC. We are making a lot of progress. You can see what the features are through laptops in other ecosystems. I think in terms of the potential of these features, this is a wake-up call for the industry. We're going to stick with it.
How will Arm win in the GPU?
When you count in units, we're actually the number one supplier of GPUs on the planet. The NVIDIA CEO made a key point. We will stay true to our credo – performance per watt counts. We're not going to risk going to make a 100-watt TDP GPU and try to make a difference in that area. We will be in the 1 watt range and try to optimize within that range. Then in some areas, you can start doing more on multiple GPUs. We're starting to see people thinking about doing machine learning extensions and possibly doing those things within GPUs. This is interesting because GPUs can benefit from some degree of AI and machine learning to paint shaders in a more efficient way. At the same time, you can start working or move ML workloads onto GPUs. There are a lot of interesting innovations that will be able to take place on our GPUs, and we're making a lot of investments in that space. To be clear, we will remain in an area where performance and efficiency are important.
You deliver a lot of GPUs. It is part of the SoC package for many customers. Customers with their own GPUs or their own GPU extensions lead the way in performance per watt. Does the same transaction not matter at the end of the day because they are shipping the Arm instruction set architecture and you can still get paid? Or is it "we have to stay competitive here or they'll leave altogether"?
GPUs are slightly different from CPUs because the API is despiration. Performance per watt is really important. We've done a lot of work for fair play and I think we're leading the way in a lot of areas. Another thing that is critical for these GPUs is the efficiency of the interface between the CPU and the GPU. This is also an area where we invest heavily, but we must be vigilant in the GPU market. No doubt. As you said, it's very competitive. You let people do it themselves, you let other third parties do the product. At the same time, this is an area that we are very focused on, so we will invest. I think the future of machine learning AI vectors used with GPUs is a huge opportunity for us.
GPUs are a huge investment. You have to spend a lot of money to compete and win the design. Are you making this investment because over time, you can win back these products from other custom GPUs and get more bang for money? Or do you just need a competitive GPU to be competitive?
Back to machine learning and artificial intelligence. If you have a heterogeneous computing system, including CPUs, GPUs, and NPUs, I think at some point the compiler will be smart enough that they might point to parts of the code that can run better on the GPU vs. CPU. If you think of the entire cluster as a complete subsystem, I think this will have long-term benefits. For us, especially when designing system-on-chip, we're trying to make all sorts of big trade-offs. Sometimes I don't want that workload to run on the CPU because it consumes unnecessary power, and I may have transistors in the GPU, and I might be able to use them if I don't draw. We believe that GPUs are a very critical strategic component.
CPU competitors are harder to identify. Who are your competitors?
I think there are probably only two options. Although biased, I don't even think there is a choice. From an ISA's perspective, you have x86. If you want to build an SoC on x86, there are only two companies in the world that can do it for you: AMD or Intel. Then there's RISC-V, which is a completely different part of the stratosphere because it's open source and there are a lot of different versions. RISC-V is a scalable processor, which means that its strengths are its weaknesses. In my opinion, extensibility leads to fragmentation. I think it hurts them to get any kind of software ecosystem. It's really hard for you to look around and say, "What's the main software ecosystem that runs on RISC-V processors continuously?" ”
Nothing. So, where do we see RISC-V in system-on-a-chip today? It is located in the deeply embedded part of the chip, and external programmers are unaware of its existence. This is an analogy I can give. If you put something in the microwave oven and then press the start key on the 30-second timer, the monitor might be Arm, because it might be running a small piece of open source application code. The timer that actually turns on the oven and closes after 30 seconds may be RISC-V. These are indeed the only two options. I mean, all of these are small companies, but these companies really don't have much appeal in terms of market share.
Do you see a world of x86s where Arm and RISC-V are completely obsolete?
x86 has been around for a long time and has a very large installation base. I certainly don't want to do anything to disparage what they're doing. I think the challenge with x86 is that it comes from only two companies. In terms of how far it can go, this in itself is a limiting factor. In fact, Intel and AMD are already very good at Arm.
Are there any other competitors like RISC-V? RISC-V is another Windows on Arm thing. It seems to appear in some of the apps you are talking about.
It's all about software. At the end of the day, it's about having a rich ecosystem of developers who can take advantage of writing software. For us, we think there are 15-25 million developers writing on Arm. Some people know they write on Arm because they take advantage of the instruction set, but others don't know because it's abstracted. Both the compiler and the open source library are there, which makes it easy to run.
Case in point: Smart TVs. If you try to write a new menu for your Smart TV and try networking. Essentially the Android version of the underlying kernel, that's Arm. It's all optimized for Arm. This is a big advantage in terms of self-reporting. These libraries are present and have been optimized. Suppose you tried it on RISC-V. If Company A adds 17 new instructions to make their RISC-V look a little different, developers won't know that. How will developers take advantage of this? RISC-V may eventually be simplified to the lowest common denominator.
A few months ago, Qualcomm CEO Cristiano Amon blamed the software developers. "Microsoft, Adobe and other large independent software vendors must join," he said. They have to make these things quickly. You just said it's all about software. Do you want to say to those companies, "Turn your attention to the future, this is Arm"?
Yes. He wasn't wrong, and in the end, that's where it was going to change. It's closer than ever, because not only is more work done on native applications, but more of these native applications have been written for Arm.
Take a look at two different operating systems in the world. Other operating systems use their names in their code. View all Microsoft applications running only on their systems. They are all ported this way, so all of these Microsoft applications can run on your phone. I think the tipping point comes when there are a lot of pretty good CPU products on the market that can compete with other people. In your opinion, this product proves that you don't have to sacrifice performance and game-changing battery life in terms of form factor.
What's next for Arm?
We will continue to invest in the areas I mentioned. We think there is really strong growth in some of the topics we talked about earlier: those complex packaging systems and complex designs. We're trying to find the ability to serve the industry more because I think there's a huge opportunity out there.
*Disclaimer: This article was created by the original author. The content of the article is his personal opinion, our reproduction is only for sharing and discussion, does not represent our approval or agreement, if there is any objection, please contact the background.