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)

According to a report from TechNews, in response to the temporary power outage caused by a severe thunderstorm on the afternoon of August 13th, memory manufacturer Nanya Technology announced on the evening of August 14th that the outage led to a halt in some machinery operations.

As emergency measures and safety checks were immediately implemented, all of Nanya Technology’s employees are safe, and the company is working to restore operations, which are expected to resume within 2-3 days.

Still, the power outage had led to damage to wafers, equipment repairs, and production interruptions, with an estimated loss of NTD 300-500 million (roughly USD 9.3 – 15.5 million). The exact figure will be confirmed after further assessment.

On the day of the power outage on August 13th, TechNews already reported that Nanya Technology’s plant experienced a 20-minute power outage during the incident. The uninterruptible power supply (UPS) was activated, minimizing damage in the photolithography and etching areas. However, sites without UPS are still being assessed for potential impacts.

On the other hand, memory giant Micron only experienced a voltage sag and reported no significant issues. Micron further stated that the voltage sag at its Taiwan facility due to the August 13th incident did not cause any harm to employees, and operations remain normal.

Read more

• [News] Taiwan Faces Temporary Power Outage in Certain Areas, while Rumors Hint at Potential Impact on Micron Equipment
• [News] Nanya Technology Chairman Names 10nm 1B Process as Key Expansion Focus for the Year

(Photo credit: Nanya Technology)

Recently, the Financial Times reported that despite over USD 400 billion in tax incentives, loans, and subsidies provided by the U.S. under the Inflation Reduction Act and the CHIPS and Science Act to boost local clean energy technology and semiconductor industries, the resurgence of U.S. manufacturing has been delayed as investors hit pause on their plans.

Reportedly, there are 114 major projects tied to these acts, totaling USD 227.9 billion in investments. However, projects with a combined investment of approximately USD 84 billion have faced delays ranging from two months to several years, with some even indefinitely postponed. These delays include several semiconductor projects.

Companies involved have cited worsening market conditions, slowing demand, and uncertainties in domestic policies as reasons for altering their investment plans.

• TSMC Delays Production at Arizona Facility

On August 13, TSMC announced several board resolutions, including the approval of a nearly USD 29,6 billion capital budget. Among these, TSMC approved up to USD 7.5 billion in funding for its wholly-owned subsidiary, TSMC Arizona.

TSMC had initially planned to build three fabs in Arizona over the next few years, with a total investment of USD 65 billion. However, per a recent New York Times report, despite four years having passed since the announcement, the Arizona plant has yet to produce a single chip.

According to a report from WeChat account DRAMeXchange, cultural differences and competition for labor resources with Intel are among the factors contributing to the challenges faced by TSMC’s Arizona facility, leading to production delays.

When TSMC announced plans to build a semiconductor fab in Arizona in May 2020, the initial plan was to start construction in 2021, with production slated to begin in 2024. The second fab was announced in December 2020, with a production target of 2026.

In May of this year, TSMC’s website indicated that the first Arizona fab’s production start has been postponed to the first half of 2025, while the second fab’s production has been delayed to 2028.

As for the third fab, TSMC has not yet disclosed the start date for construction, but the official plan is to commence production by the late 2030s.

According to TSMC’s plan, the first Arizona fab will use 4nm process technology, the second fab will employ 2nm technology, and the third fab will utilize 2nm or more advanced process technologies.

• Intel Delays Construction of Ohio Project

Intel, the U.S. semiconductor manufacturer, plans to invest USD 100 billion over the next five years in new fabs and expansions across Arizona, New Mexico, Ohio, and Oregon, creating 10,000 manufacturing jobs and 20,000 construction jobs.

Yet, according to a previous report by The Wall Street Journal in February, Intel has delayed its USD 20 billion chip project in Ohio due to market downturns and delays in U.S. subsidies.

Intel is set to build two new advanced fabs in Ohio, with an initial plan to begin chip manufacturing in 2025. Following adjustments, the completion of Intel’s Fab1 and Fab2 projects in Ohio has been postponed to 2026–2027, with operations expected to commence around 2027–2028.

As chip manufacturing processes advance to 3nm and 2nm, the investment required for fabs has surged, putting semiconductor companies under financial pressure. Against this backdrop, Intel has not only delayed the construction of its Ohio facility but has also made adjustments to its European projects.

Intel’s planned EUR 30 billion investment in two fabs, Fab 29.1 and Fab 29.2, in Magdeburg, Germany, was initially set to start in the second half of 2023.

However, due to delays in confirming EU subsidies and the need to remove topsoil at the construction site, Intel has postponed the start date to May 2025. Additionally, Intel has also paused its investment plans for facilities in France and Italy.

• Samsung Delays Taylor Fab Project

Initially, Samsung planned to build a semiconductor cluster in Taylor, Texas, including two advanced logic fabs and one advanced packaging facility, with up to USD 6.4 billion in U.S. subsidies.

The first of these fabs in Taylor began construction in 2022, initially scheduled to start production in 2024 with 4nm process capabilities. However, the plant may not begin operations until 2026, US local media MySA noted. This delay is likely due to a slowdown in the foundry market and delays in the disbursement of U.S. subsidies.

Meanwhile, according to reports from Tom’s Hardware and the Korean media outlet ETnews, with the delay in the construction of the semiconductor plant, Samsung may upgrade the facility’s advanced process technology from 4nm to 2nm.

This adjustment aims to enhance Samsung’s competitive edge in advanced process, positioning it more effectively against rivals like TSMC, Intel, and Rapidus.

Read more

• [News] Nearly 40% of Major Manufacturing Projects Reportedly Face Delays under U.S. IRA and CHIPS Act
• [News] Opening of TSMC Kumamoto Plant Nears, Yet Delay in Arizona Plant – Why is US Semiconductor Fab Construction Lagging Globally?

(Photo credit: TSMC)

Samsung Electronics, which has been struggling at the final stage of its HBM3e qualification with NVIDIA, may unexpectedly emerge as the pacemaker for the AI ecosystem, as the company may somehow ease the cost pressure for building AI servers by balancing the market, as well as alleviating the tight HBM supply, according to a recent report by Korean media outlet Invest Chosun.

Samsung, in its second quarter earnings call, has confirmed that the company’s fifth-generation 8-layer HBM3e is undergoing customer valuation. The product is reportedly to enter mass production as soon as the third quarter.

Invest Chosun analyzes that while there is growing anticipation that NVIDIA could come up with a conclusion regarding Samsung’s HBM3e verification, the market’s attitude towards AI has also been gradually shifting in the meantime, as the main concern now is that semiconductors are becoming too expensive.

The report, citing remarks from a consultant, notes that the price of an NVIDIA chip may cost tens of thousands of dollars each, leading to concerns that the industry’s overall investment capex cycle might not last more than three years.

In addition, the report highlights that the cost of building an AI server for learning is about 40 times that of a standard server, with over 80% attributed to NVIDIA’s AI accelerators. Due to the cost pressure, big techs have been closely examining the cost structure for building AI servers.

Therefore, NVIDIA has to take its customers’ budgets into consideration when planning its roadmap. The move has also sparked speculation that NVIDIA, which is prompted to lower product prices, might compromise to bring Samsung onboard as an HBM3e supplier, the report states.

Citing an industry insider, the report highlights the dilemma of NVIDIA and its HBM suppliers. As the AI giant tries to shorten its product cycle, releasing the Blackwell (B100) series just two years after the Hopper (H100), HBM suppliers have been struggling except for SK hynix, as the company is the only one with the most experience.

If Samsung doesn’t join the HBM lineup, the overall supply of NVIDIA’s AI accelerators could be limited, driving prices even higher, the report suggests.

Under this backdrop, Samsung may have taken on the role of pacemaker in the AI semiconductor market, as it may help balance the market during a time when there are concerns about overheating in the AI industry. Also, if it is able to form a strong collaboration with NVIDIA by supplying 8-layer HBM3e, its technological gap with competitors will noticeably narrow.

TrendForce notes that Samsung’s recent progress on HBM3e qualification seems to be solid, and we can soon expect both 8hi and 12hi to be qualified in the near future. The company is eager to gain higher HBM market share from SK hynix so its 1alpha capacity has reserved for HBM3e. TrendForce believes that Samsung is going to be a very important supplier on HBM category.

Read more

• [News] Samsung’s 8-layer HBM3e to Start Mass Production in Q3, Driving HBM Sales to Soar 3-5 Times in 2H24
• [News] Battle between Memory Giants Heats up in 2H24 as Samsung and SK hynix Advance in HBM3/ HBM3e

(Photo credit: Samsung)

After announcing reductions in capital expenditures, massive layoffs, and the suspension of dividends, according to a report from Reuters, Intel is focusing on increasing its cash reserves to sustain company operations. The same report also indicates that Intel has sold off its entire stake in the intellectual property company Arm.

Per another report by wccftech, based on Intel’s Form 13-F filed with the U.S. Securities and Exchange Commission (SEC), the company has disposed of its entire stake in Arm Holdings, totaling 1.18 million shares. Reuters further reported that through this sale, Intel would have raised approximately USD 146.7 million.

Despite selling its shares in Arm, Intel has retained its holdings in companies like Astera Labs, Joby Aviation, MariaDB, and Senti Biosciences. Yet, as per the same report from wccftech, these investments have yet yielded significant returns, with Intel currently experiencing a cumulative loss of USD 120 million on them in Q2.

At that time, Intel CEO Pat Gelsinger pointed out that Intel must align its cost structure with the latest operational model and fundamentally change the way the company operates. He indicated that Intel’s revenue growth has not met expectations and has not yet benefited from powerful trends such as AI.

Read more

• [News] Intel’s Earnings Fall Short, Cutting Over 15% of Workforce
• [News] Intel Suspended Investment in French and Italian Chip Plants

(Photo credit: Intel)