Taiwan’s Tech Sector: Adapt or Die
Quanta Forging a New Future
Software has taken over the global electronics sector, leaving Taiwan’s hardware makers in a bind. But Quanta Computer has embraced a new path that builds on its past, as Quanta Chairman Barry Lam explains.
Quanta Forging a New FutureBy Liang-rong Chen, Elaine Huang
From CommonWealth Magazine (vol. 626 )
After years of feeling their way and learning through constant trial and error, Taiwan’s notebook computer suppliers may finally be seeing a glimmer of light in their efforts to reinvent themselves.
Quanta Computer has been quietly cultivating the American market for years, preparing to provide solutions to hyperscale data center operators through “private cloud” brand QCT (Quanta Cloud Technology), and the initiative is starting to bear fruit.
In fact, the cloud business could contribute as much as half of Quanta’s total profits this year.
The company is the world’s largest contract notebook computer manufacturer, but with the notebook business’s best days behind it, Quanta looked elsewhere for growth and successfully reinvented itself as a bona fide cloud technology services company.
It’s no wonder then that Quanta’s market capitalization of nearly NT$280 billion roughly equals that of Taiwan’s other major notebook makers – Compal Electronics, Wistron and Inventec – combined. Of the four heavyweights dominating the fading notebook sector, one has clearly pulled away.
Marginalized by Its ‘Hardware’ Past?
A few key aspects of Quanta’s transformation particularly stand out.
First, the global consumer electronics market has been dominated by Apple in recent years, and the only Taiwanese electronics contractors who have made money are Apple suppliers, such as iPhone assembler Hon Hai Precision Industry (Foxconn) and MacBook assembler Quanta.
Now that Apple’s growth has started to plateau, many people wonder what will happen to Taiwan’s tech sector when Apple’s aura evaporates.
Quanta and Hon Hai have both embarked on journeys that have made them models of transformation. Hon Hai acquired Japan’s Sharp Corporation to expand its scope and reach. Quanta used some of its profits from the Apple business to invest in software development and build a manufacturing facility and service center in Silicon Valley to create its own brand.
The second significant aspect of Quanta’s transformation is that QCT, the private cloud brand created by Quanta Chairman Barry Lam, is an industrial brand. Its business model of working with corporate clients and tailoring solutions to their needs resembles the strategies of two other venerable Taiwanese computer companies that have adapted well to changing times – Advantech Corporation and Delta Electronics.
B2B branding is anything but glamourous and laying the groundwork takes a long time, but it offers relative stability compared to the volatile existence of consumer tech brands. Well-known Japanese electronics companies have jumped on the bandwagon, with NEC and Panasonic shifting their resources into the B2B realm, hoping to capitalize on booming emerging market demand for infrastructure and for corporate services.
Under the leadership of Quanta, Advantech and Delta Electronics, the strategy of forging industrial brands in the information technology sector can perhaps become the new road to salvation for Taiwan’s tech companies.
The final, and most important, aspect of Quanta’s transformation has been the relatively new model of hardware-software integration.
In the past few years, the tech sector’s spotlight has been fixated on software services. America’s four tech giants – Google, Facebook, Amazon and Apple – not only make much higher profits than high-tech hardware companies, they have also delved into hardware fields, making, for example, high-performance ICs and VR headsets.
The dominance of software suggests that maybe Netscape founder Marc Andreessen was right when he wrote in a now famous essay in the Wall Street Journal in 2011 that software is eating the world.
Amid this huge paradigm shift, could Taiwan, founded on its hardware prowess, be completely marginalized by the new economy?
Quanta’s new model is the answer given by Barry Lam on behalf of Taiwan’s electronics manufacturing sector.
Software + Hardware = The Best User Experience
“The new economy is a service-based economy. Taiwan’s weakness is that it doesn’t have that big of a market, so the United States and China dominate. Along with that, Taiwan’s software industry is a laggard. Countries that don’t have big markets are not likely to have powerful software sectors. For us to try to use our weakness to compete with the strength of others would be very difficult,” Lam says bluntly.
In new internet fields involving purely software, Taiwan is inherently lacking. But in pure hardware manufacturing in which Taiwan excels, value is being eroded by the day.
Under Quanta’s new “private cloud brand” model, it makes the servers, storage systems and switches, and hardworking Taiwanese software engineers integrate that with proprietary software or free software to provide complete solutions for companies or organizations in the financial, old economy and medical sectors.
The key to the hardware-software integration model is that while the core value of the service lies in the embedded software, it still has to be bundled and used together with a hardware package to achieve the best possible “user experience.”
In Barry Lam’s eyes, that model has the potential to liberate Taiwan’s tech sector from the clutches of the contracting model spawned when Intel and Microsoft (Wintel) ruled the industry.
In an interview with CommonWealth Magazine, Lam sees that transition and a move toward cloud and artificial intelligence applications as giving Taiwan new impetus for innovation. Excerpts of the interview can be found below.
CommonWealth Magazine: Many people think Quanta just makes servers on an OEM basis. But what they don’t know is that you have developed a new business. What is it?
Barry Lam: We are now in the solutions business, rather than just making servers. IBM is like that. When it started, it made computers but later started providing solutions.
In the past, we were only at the manufacturing end of the value chain. In this field, it used to be Intel on this side and Microsoft on that side, and we were sandwiched in between. But that’s not the way it is anymore.
In this new environment, we have a lot of maneuvering room. There used to be just one application – Windows – and we had no flexibility to make any changes. But today, it’s different. In cloud computing and AI, there are a lot of applications.
The scientist Stephen Hawking has said there are three things that will change the world: climate change, biotechnology and artificial intelligence.
AI has a lot of applications, and we absolutely have to be involved. One example is AI cloud computing. We have a team of people learning about AI technology and applications. They are studying how to make AI equipment better and looking for ways to give AI applications more depth so that our solutions can be more complete. This has picked up last year and this year, not only server manufacturing but also cloud solutions.
CW: Why have things picked up?
Lam: Because when the market had the need, we had the product ready. Customers won’t wait for you.
CW: How did you foresee the trend?
Lam: We collaborated with MIT about four or five years ago. MIT was working on AI, and I thought that was a big deal.
From a computer science perspective, AI is a big deal. When I went (to MIT) in 2000, even then I saw it as something important and asked their Computer Science and Artificial Intelligence Laboratory to help.
It’s like when the iPhone appeared 10 years ago, the internet economy exploded. In the next 10 years, I believe AI will be a new economy.
Small Clouds the Trend
CW: What progress has Quanta made in cloud computing in response to new technology trends?
Lam: In the past, the smaller you could make a server, the more you could fit in a rack, and that would give us the edge over others.
But today, that isn’t important anymore. What’s important now is what applications does it have? What applications can you develop based on the servers that I supply to you? We integrate servers, storage systems and switches together to create a small cloud.
We used to be involved in the big cloud, but in the AI era, small clouds are the trend. Only with different combinations can computing be more effective.
In the future, AI, robots and driverless cars – all of these use GPU servers. Whatever related processor Nvidia or Intel makes, we will integrate it into our solution and give customers different choices based on their needs. We have spent a lot of time on this and researched it in great depth.
There are no more than 10 providers involved in the big cloud around the world. They are internet service providers and work with huge customers in the United States and China.
We’ve been very active in the AI cloud, and demand for small clouds will be especially evident in the 5G era. Why small? Because they have to be close to the user. How clouds are connected will depend on 5G, and this will be called “distributed clouds.” These distributed clouds will rely on 5G to communicate with each other, because only 5G is fast enough to satisfy the needs of AI applications.
Facebook has had an “open cloud project” in which all of the racks’ specifications are standardized. But with AI’s distributed clouds, every cloud has different specs, because they put a premium on specialization.
CW: Is making “distributed clouds” a good business for Quanta?
Lam: If we’re able to do it, it’s a good business. If we’re not, it’s not a good business. (Laughs) But of course to be able to do this, the skills you have to develop are a little more sophisticated.
CW: How so?
Lam: Before it was like hunting an elephant. Now it’s like hunting a tiger. You have to move faster, you have to get closer to the customer and closer to the applications.
CW: How will the process affect Quanta internally and what changes need to be made?
Lam: We have to innovate and we have to take the initiative. We can’t be passive. To give you an example, contract manufacturing is like a self-serve cafeteria. I make a certain amount of food, and you eat it. But this is like a gourmet restaurant, where each dish has to be special. Self-serve cafeterias are all alike, but gourmet restaurants are distinct from each other.
CW: What are some of your secrets for building a unique, gourmet restaurant?
Lam: We have to keep developing new technologies and exploring and talking to customers about applications. There’s also marketing services. So we set up the QCT cloud solutions center in the U.S.
If all we were running were self-serve cafeterias, we wouldn’t have to do this. If a customer wanted something, we could just fly over there. But cloud computing is different; it’s about having our customers experience [a solution], even testing solutions they want in front of them so they can decide if it’s the best solution for them.
When we started making notebooks, we also took the initiative to suggest specs, and if customers liked them, they would recognize the value we provided. It’s the same idea with our servers today. We develop and make them and then apply for patents; otherwise you end up always being at the mercy of others.
CW: In what areas do you plan on investing the most in the future?
Lam: Now that’s a secret. (Laughs) But AI, driverless cars, future medical care – these are the big directions. In terms of applications, medical care is something everybody has to use.
Right now, 5G is a big change, and it will be commercialized by 2020. We’re in contact with a few 5G companies to talk about what kinds of clouds we could create that they would like.
In terms of clouds, we’re the biggest company and our R&D is the strongest. Why wouldn’t they want to work with us? Only if you take the lead in a field can you have that kind of advantage.
CW: What kinds of 5G companies are you referring to?
Lam: Lots of them. (Laughs) We believe that 5G and AI represent the rise of a major era.
It’s like in 2000 when we invested in servers, we didn’t know that Google or Facebook would emerge on the scene, but they had to buy hardware, and seeing what they bought, you knew that a new trend and opportunity had arrived.
[The model] is a lot different from the past, but this way is more interesting and offers more room for profits. Our engineers are very excited because they get to do something different.
CW: Do you need to recruit new people for the AI era?
Lam: We’ve brought in a group of AI Ph.Ds, but Taiwan hasn’t really started yet on AI applications.
CW: You seem to be doing a lot of cloud business now. How come its contribution to gross margin hasn’t become evident yet?
Lam: We have many areas that are new that have higher margins, but their R&D costs are also high. Right now, we are investing heavily, but our turnover has yet to surge.
It looks like it will pick up in 2020. We spent a lot of money on our office in the United States and hired a lot of people and only now are we starting to recoup some of that, because it usually takes two years of steady contact with a customer before you get results.
There won’t be any sudden and explosive growth. There are no free lunches in the world. I’ve never won a lottery my entire life. (Laughs) In the industrial sector, there just aren’t any opportunities for reaching the sky in a single bound.
CW: You previously described Quanta as a tortoise. How would you describe today’s Quanta?
Lam: We still operate one step at a time, like a tortoise, but now we raise our heads up a little higher.
Translated from the Chinese by Luke Sabatier