According to TrendForce’s “2024 Global GaN Power Device Market Analysis Report”, the development of the GaN power device industry is expected to accelerate once again as Infineon and Texas Instruments allocate more resources into GaN technology.

In 2023, the market size of global GaN power device was around USD 271 million, and it is projected to grow to USD 4.376 billion by 2030 at a compound annual growth rate (CAGR) of 49%.

Notably, the proportion of non-consumer applications is expected to increase from 23% in 2023 to 48% by 2030, with automobile, data center, and motor drive being the core application scenarios.

• GaN Expected to Make Great Difference in Reducing Heat and Improving Efficiency in AI Era

The evolution of AI technology has driven the continuous increase in computing power demand, making the power consumption of CPU and GPU an increasingly striking issue. To meet the requirements of more advanced AI computations, server power supply is required to further enhance efficiency and power density, and thus, GaN has emerged as a key solution.

Delta, the world’s largest server power supply provider, holds nearly 50% of the market share. Observing the advancement of its server power supplies, the power density has increased from 33.7W/in³ to 100.3W/in³ over the past decade, while power levels has reached 3.2kW and even 5.5kW, and the next generation is expected to exceed 8kW.

TrendForce’s research indicates that AI server is expected to account for 12.2% of overall server shipment in 2024, an increase of ~3.4% from 2023, while the annual growth rate for general server shipment is only 1.9%.

In face of such an attractive opportunity, both Infineon and Navitas Semiconductor have announced technical roadmaps for AI data center this year.

Infineon highlights the significant advantages of combining liquid cooling technology with GaN at lower junction temperature, which will enable data center to maximize efficiency, meet the growing power demands, and overcome the challenges posed by server heat increase.

• Potential of GaN’s High-Frequency Characteristics can be Fully Tapped in Motor Drive Applications

In motor drive applications like robotics, the potential of GaN is gradually emerging. Compared to industrial robots, humanoid robots have a significantly higher degree of freedom (DoF), greatly increasing the demand for motor drivers.

It’s learned that the joint modules of humanoid robots bear the main tasks of exertion and braking. To achieve higher explosive power, motor drivers with high power density, high efficiency, and high responsiveness are needed. As a result, GaN has attracted market attention, especially in load-bearing areas like the legs.

Texas Instruments and EPC (Efficient Power Conversion) have been dedicated to driving GaN’s application in the motor drive field, drawing new players into the market.

Robotics is expected to embrace a future beyond imagination, where precise, fast, and powerful motion capabilities are crucial, and the motors driving these movements will inevitably advance forward, which will be a boon for GaN.

• GaN Provides New Solutions for Automotive Power Electronics

While SiC thrives in the automotive industry, GaN is also gaining traction in this field, with on-board chargers (OBC) considered the best entry point.

The first automotive-grade GaN power product meeting AEC-Q101 standard was released by Transphorm (now Renesas) in 2017, and several manufacturers have since introduced a wide range of automotive-grade products so far.

Overall, although GaN still faces several technical challenges in entering inverter and OBC power system, it is believed that with continuous investment from major automotive chip companies like Infineon and Renesas, GaN will soon become a key component in automotive power systems.

• Consumer Electronics Remain the Main Applications for GaN

Consumer Electronics still holds the biggest proportion among GaN power device applications, in which GaN’s footprint is quickly expanding from fast chargers to home appliances and smartphones.

Specifically, GaN has been widely adopted in low-power smartphone fast chargers, and next will enter into more demanding applications like notebook and home appliance power supplies. Other potential consumer applications include Class-D audio, smartphone over-voltage protection (OVP), etc.

TrendForce believes that GaN power device industry is at a critical breakthrough moment, with several potential applications simultaneously boosting rapid growth.

Moreover, new structures and processes are expected to be introduced in built on better reliability to get into more complex high-power, high-frequency scenarios, injecting new momentum into the industry.

In terms of industry development and market landscape, Fabless companies have been particularly active in the past.

However, as the industry continues to consolidate and the application markets gradually open up, traditional IDM (integrated device manufacturer) giants are expected to gain significant influence, bringing new major changes to the future landscape of the industry.

Read more

• [News] Foundry Giant GlobalFoundries Acquired GaN-Related IP
• [News] SK keyfoundry Advances in GaN Power Semiconductors, Reportedly Producing for Tesla Soon

(Photo credit: Infineon)

After the disappointing financial performance in Q2, Intel has been plagued by a series of challenges. Its shares plunged, falling to the lowest point in over a decade, coupling with the suspension of dividend payouts, layoffs, and product failures. The company may be facing its worst moment in 50 years, and now, the market seems to lose patience with the once-dominant semiconductor giant.

Even Pat Gelsinger, who took over as CEO in 2021 with his technical expertise, has yet been able to rescue this struggling mammoth.

Over the past 50 years, Intel has witnessed eight CEOs, as each of them has left indelible marks in history. Some of their decisions have made Intel so sluggish that it is difficult to turn around.

The latest report by Technews looks back at Intel’s past and summarizes its former CEOs’ impact, examining how the tech giant’s development trajectory is closely intertwined with that of semiconductors.

What happened before Intel’s Foundation?

Let’s go back to 1956, when William Shockley, known as the ‘Father of the Transistor,’ left Bell Labs to establish the Shockley Semiconductor Laboratory.

However, Shockley’s authoritarian management style, as well as his erratic behavior, led to the dissatisfaction of employees. Eight employees, called by Shockley as the “Traitorous Eight,” resigned in 1957 to found Fairchild Semiconductor.

Fairchild, though developed rapidly, faced organizational management issues as well. In July 1968, two of its founders, Robert Noyce and Gordon Moore, resigned to found Integrated Electronics, later known as Intel, on July 18th of the same year. Andy Grove joined as the third employee afterwards.

They are regarded as Intel’s founders who formed a management triad. Noyce was in charge of research and development. Moore, on the other hand, was in charge of business execution, while Grove responsible for commercialization and management. The Big Three had led Intel through its first 30 years, establishing the glorious era from 1968 to 1998.

(Left to right:  Gordon Moore, Robert Noyce and Andy Grove)

First CEO: Robert Noyce (1968–1975)

During Noyce’s reign, Intel introduced the first microprocessor in 1971, which marked the beginning of the personal computer (PC) era and the start of Silicon Valley’s golden age. The breakthrough has earned Noyce the nickname of “Mayor of Silicon Valley” or “Father of Silicon Valley.”

It’s worth noting that Noyce and Texas Instruments engineer Jack Kilby were both pioneers of the integrated circuit (IC). Kilby developed the world’s first integrated circuit at Texas Instruments in 1958, earning him the title of “Father of the IC.” The following year, Noyce conceived the concept of the first planar IC with a metal interconnection method in his notebook, laying the foundation for all modern IC technologies.

Second CEO: Gordon Moore (1975–1987)

The famous “Moore’s Law” was proposed by Moore in 1965. However, the concept was somewhat speculative, as the development of integrated circuit was still in its early stages.

Moore’s prediction was aimed to convey the idea that electronic products would become increasingly affordable, and surprisingly, it came true. Considering the exponential growth in IC complexity, Moore revised his forecast in 1975, stating that the number of transistors on an IC would double approximately every two years. The law has established the foundation for the semiconductor industry. Even how, it is still a hot topic today among semiconductor giants such as TSMC, Intel and NVIDIA.

Third CEO: Andrew Grove (1987–1998)

To follow his superior Moore, Grove left Fairchild to become Intel’s third employee, and took up the management duties after Noyce and Moore.

In the 1970s, Intel’s main products were DRAM and SRAM. As Japanese companies began to flood the global market with DRAM, the profit of the product line quickly declined. Therefore, Grove decided to discontinue DRAM-related products and focus on integrated circuit applications instead.

The decision helped Intel to seize the vast opportunities of the PC era.

Additionally, one of the critical decisions during his tenure was to manufacture its 386 processor independently, which successfully showcased Intel’s capability to manufacture its own processors, establishing its undisputed leading position in the early 1990s.

When Intel was founded, its annual revenue was only USD 2,672. By 1997, thirty years later, its annual revenue had grown to USD 20.8 billion. Grove played a crucial role in the success and was named by Time magazine as Person of the Year in 1997. He also documented his management philosophy in his book, with the famously saying, “Only the Paranoid Survive.”

Fourth CEO: Craig Barrett (1998–2005)

Craig Barrett was an associate professor of Materials Science and Engineering at Stanford University before joining Intel. Upon taking office, his primary challenge was determining whether Intel could become a company that could handle “low margins.”

At that time, the market believed the high-profit era of the semiconductor industry was over, and that the future of PCs would be dominated by low-cost models. Therefore, he led Intel through two major transformations.

The first transformation was the segmentation of Intel’s processor products. Due to the company’s rapid transformation, the competitors are difficult to follow, making them unable to disrupt Intel’s position in the low-cost market. The second, on the other hand, was the expansion from computer/ computing into network servers.

Barrett also believed that Intel’s competitiveness lied in manufacturing and R&D, so he invested USD 28 billion in building advanced facilities and developing new technologies, which secured Intel’s leadership in manufacturing technology.

Fifth CEO: Paul Otellini (2005–2013)

Under Otellini’s leadership, Intel underwent another significant shift. As the first CEO in Intel’s history without an engineering background, Otellini only held an MBA degree.

During his tenure, Intel’s financial performance was excellent, but the company’s focus had been moved from technology to performance-oriented, which prioritized sales and marketing over technological advancements. This shift set the tone for Intel’s later decline.

In 2005, Intel secured an order from Apple, which would adopt Intel’s chips in Macs. However, when Apple inquired if Intel would supply processors for iPhones, Otellini declined the request as he believed the deal was not cost-effective. This decision caused Intel to miss out on the booming mobile device market following iPhone’s 2007 launch.

Rather than focusing selling chips individually, Otellini believed Intel’s platform had greater value, which helped Intel secure its share in the x86 market. However, the company’s technological advantage began to wane. Moreover, due to the global economic downturn, Intel closed five factories, including its last plant in Silicon Valley.

Sixth CEO: Brian Krzanich (2013–2018)

With the decline in the PC market, the new CEO Brian Krzanich, responsible for technology and administration, faced the critical task of transformation once again. In his tenure, Intel shifted its focus towards the Internet of Things (IoT) and cloud computing.

However, as Krzanich did not believe in the economic scalability of EUV, he opted to forgo ASML’s first-generation EUV equipment. As a result, Intel’s 10nm progress faced multiple delays, causing it to fall behind competitors like TSMC and Samsung in advanced nodes, and even led to a loss of market share to rival AMD.

Seventh CEO: Robert (Bob) Swan (2019–2021)

Intel was plagued with many issues, such as the problems with the 10nm process, which were difficult to overcome. During Swan’s tenure, Intel’s dominance in the market has gradually declined. In some sectors, AMD even caught up and overtook Intel’s throne.

At the same time, instead of using Intel’s chips, Apple introduced its self-designed M1 processors, which might become the final straw to Intel’s dominance.

Additionally, Intel had discussions with OpenAI about investment opportunities in 2017–2018, but Bob Swan believed that generative AI models would be challenging to commercialize in the short term and considered the deal to be unprofitable, thus giving up the opportunity to participate in the AI boom.

Eighth CEO: Pat Gelsinger (2021–)

As the three former CEOs all came from operations or finance backgrounds, Pat Gelsinger, with his technical background, was expected to bring a fresh outlook to Intel.

Given the challenge ahead, he planned to significantly expand Intel’s factories and announced the “Four Nodes in Five Years” plan, aiming to advance five nodes within four years, betting Intel’s future on the 18A process.

However, before the 18A technology was introduced, Intel has already been plagued by layoffs, suspension of dividends, and a stock price nearing tangible book value. Whether Gelsinger is the savior to lead Intel out of the woods remains to be seen. Only time will tell.

Read more

• [News] Intel Reportedly Rejected OpenAI Investment, Missing Out on AI Opportunity
• [News] Intel Delays Innovation 2024 Event amid Financial Challenges, Arousing Public’s Curiosity on 18A

(Photo credit: Intel)

According to a report from the Financial Times, SoftBank Group has decided to move away from its plan to collaborate with Intel on producing AI chips to compete with NVIDIA and is now reportedly focusing on discussions with TSMC.

The same report, citing sources, reported that the partnership between SoftBank and Intel fell through because Intel struggled to meet SoftBank’s requirements. SoftBank reportedly attributed the collapsed talk to Intel’s inability to meet their demands for production volume and speed.

The report noted as well that this fallout occurred before Intel’s announcement of releasing its official announcement on its Q2 (April-June) earnings in early August. Notably, in response to a significant drop in its performance, Intel planned to lay off about 15,000 employees and suspend shareholder dividends.

Moreover, the report further cited rumors claiming that SoftBank has shifted its focus to discussions with TSMC; however, no agreement has been reached so far.

Reportedly, Intel, SoftBank and TSMC have all declined to comment on the situation.

Read more

• [News] Intel Has Offloaded Stake in ARM, Reportedly Raising nearly USD 150 Million
• [News] Intel’s Earnings Fall Short, Cutting Over 15% of Workforce

(Photo credit: Intel)

Google has accelerate its pace on the Pixel series, as the tech giant launched Google Pixel 9 on August 13th, which is two months ahead of its schedule.

Though the Tensor G4 processor in the model is manufactured with Samsung’s 4nm, according to a report citing sources by Commercial Times, Google is said to be switching to TSMC’s 3nm process with its next-generation Tensor G5, coupling with the foundry giant’s InFO-POP packaging.

Google’s Pixel 8 is said to be the first AI-centric smartphone, featuring a range of AI functionalities. Yet, Commercial Times’ report has indicated that, after years of close collaboration, Google will part ways with Samsung and have TSMC produce the Tensor G5 chip.

The chip is also said to adopt TSMC’s advanced InFO-POP packaging. Google’s move, according to the report, demonstrates its ambition to expand its leadership in software to hardware, as it eyes for the opportunities of edge AI.

Industry sources cited by the report further point out that in the fourth quarter, both Qualcomm and MediaTek will launch flagship-level chips, while Apple’s A18 will also be produced using TSMC’s N3 process.

All these developments have hinted at tech giants’ ambition on the massive potential of the edge AI market. Now, Google would be the latest competitor to join the race.

Meanwhile, though Pixel’s market share is relatively low, the Android ecosystem, with its 70% market share in smartphones and billions of users, offers significant potential. Google is said to be following a path similar to Apple’s, achieving complete integration of hardware and software to maximize this potential.

Google’s self-developed chip extends beyond mobile devices, with its TPU (Tensor Processing Unit) now in their seventh generation. Additionally, Google’s Arm-based CPUs are being developed in partnership with TSMC.

Read more

• [News] Breaking Apple’s Monopoly – TSMC’s InFO Packaging Reportedly Adds Google Chips
• [News] TSMC Pushes for FOPLP Mass Production by 2027, Reportedly Eyeing on Innolux’s Plant

(Photo credit: Google)

At its earnings call on August 14th, Taiwanese tech giant Foxconn addressed the progress of the highly anticipated launch of NVIDIA’s next-generation AI server, the GB200.

According to a report by Commercial Times, Foxconn confirmed that the development of the server cabinets is on schedule, and the company will be among the first suppliers to deliver, with shipments expected to begin in the fourth quarter. The company also expects significant growth throughout the year, driven by strong demand for AI servers.

Addressing concerns about the progress of its AI server business, Foxconn spokesperson James Wu stated that driven by strong customer demand, the company’s AI server revenue grew by more than 60% quarter-over-quarter in Q2, accounting for over 40% of its total server revenue.

Wu reaffirmed that the robust demand for AI servers is expected to continue, and the company maintains its forecast that AI servers will contribute 40% of its total server revenue for the year.

Additionally, Foxconn has observed strong demand from various types of customers for the new generation of AI cabinet solutions, which is expected to significantly contribute to its server revenue in 2025.

Previously, per a report from The Information, NVIDIA’s GB200 is said to be experiencing yield issues, leading to a one-quarter delay in mass shipments. When asked about potential delays in the shipment of GB200 AI servers, Wu responded that the development timeline for the GB200 cabinets is on schedule.

In the second quarter, Foxconn reported revenue of NTD 1.55 trillion (roughly USD 48 billion), with a gross margin of 6.37% and earnings per share (EPS) of NTD 2.53.

For the first half of the year, its revenue reached NTD 2.87 trillion (roughly USD 88.9 billion), marking a 4% year-over-year increase. Gross margin, operating margin, and net profit margin were 6.37%, 2.83%, and 1.98%, respectively, all showing improvements compared to the same period last year, which stood at 6.21%, 2.58%, and 1.66%.

Cumulative EPS for the first half was NTD 4.12. Looking ahead to the third quarter, Foxconn expects operations to gain momentum as they enter the traditional peak season, with both quarter-over-quarter and year-over-year growth anticipated.

Read more

• [News] Foxconn Remains Unfazed Despite NVIDIA’s GB200 Delay, with Customers Aim to Secure Their Orders
• [News] GB200 AI Servers Generate USD 210 Billion Annual Revenue, Boosting Supply Chain Growth

(Photo credit: Foxconn)