Is the winter really coming for the memory sector? Despite an earlier report by Morgan Stanley warning of an AI bubble, U.S. memory giant Micron reveals a financial guidance that beats market expectations, projecting its fiscal first-quarter revenue to reach USD 8.7 billion, higher than an average analyst estimate of USD 8.32 billion, Bloomberg notes.

Meanwhile, Micron expects a significant increase in gross margin to around 39.5%, and an adjusted earnings of USD 1.74 per share, exceeding analysts’ estimates of USD 1.65, according to Reuters.

The growth momentum will mainly rely on the soaring demand for HBM, driven by AI. Earlier in June, Micron noted that its HBM chips have been fully booked for 2024 and 2025.

In terms of the outlook for the overall HBM market, Micron’s view evidently contradicts with that of Morgan Stanley, as it eyes the HBM total available market (TAM) to grow from approximately USD 4 billion in 2023 to over USD 25 billion in 2025.

And the company is making strides in its progress in HBM in the following year. According to its press release, Micron expects its HBM, high-capacity D5 and LP5 solutions, and data center SSD products to deliver multiple billions of dollars in revenue in fiscal 2025.

The U.S. memory giant also expects its HBM market share to commensurate with the company’s overall DRAM market share sometime in 2025.

According to TrendForce, Micron ranked third in DRAM revenue in Q2, 2024, with a market share of 19.6%, after Samsung’s 42.9% and SK hynix’s 34.5%, respectively.

Regarding the latest development on HBM, after its 8-hi HBM3E entered mass production in February, Micron confirms that it has started shipments of production-capable HBM3E 12-hi 36GB units to key industry partners to enable qualifications across the AI ecosystem, stating that its HBM3E 12-hi 36GB delivers 20% lower power consumption than its competitors’ HBM3E 8-hi 24GB solutions while providing 50% higher DRAM capacity.

The company expects to ramp its 12-hi HBM3E in early 2025 and increase the 12-hi mix in the overall shipments throughout the year.

According to a previous report by Tom’s Hardware, the new products are reportedly designed for cutting-edge processors used in AI and high-performance computing (HPC) workloads, including NVIDIA’s H200 and B100/B200 GPUs.

Micron delivered a strong finish to fiscal year 2024, with fiscal Q4 revenue at the high end of its guidance range and gross margins and earnings per share (EPS) above the high end of its guidance ranges.

In fiscal Q4, Micron’s revenue jumped 93% YoY to USD 7.75 billion. Its earnings per share (EPS) came in at USD 1.18, a notable turnaround from the loss of USD 1.07 per share in the same period of 2023. In addition, it achieved record-high revenues in NAND and in its storage business unit.

Micron’s fiscal 2024 revenue grew over 60%, with company gross margins expanding by over 30 percentage points and achieved revenue records in data center and in automotive, according to its press release.

Read more

• [News] Micron’s 12-Hi HBM3e Ready for Production, Targeting NVIDIA’s H200 and B100/ B200 GPUs
• [News] Morgan Stanley Predicts ‘DRAM Winter,’ But TrendForce Remains Less Pessimistic

(Photo credit: Micron)

Earlier this month, reports emerged that Samsung Electronics was cutting up to 30% of its overseas workforce, with Business Korea noting issues with its 2nm yield and the company’s decision to pull staff from its Taylor plant. Now, Digital Daily reports that Samsung has further delayed construction and orders for its Pyeongtaek Phase 4 (P4) facility and the second foundry plant in Taylor, Texas.

Citing semiconductor industry sources, Digital Daily revealed that Samsung recently notified key suppliers about the postponement of equipment and infrastructure orders for both facilities. While Samsung has not officially confirmed the construction timeline, insiders say both plants were originally slated to break ground in the second half of this year. The latest notifications, however, confirm delays.

Market speculation surrounding these projects has been rife. Digital Daily reported that some predict part of the P4 line, initially intended for foundry production, might be converted into memory production. Others suggest the expansion could be delayed altogether due to the downturn in memory semiconductor demand.

According to Digital Daily, the Pyeongtaek P4 line—designed as the world’s largest semiconductor production facility—was expected to handle both memory and foundry production. The facility is divided into four phases, with the first already operational. However, construction on phases 2-4, initially scheduled for this year, has been postponed, delaying related equipment and infrastructure orders as well.

Samsung’s foundry project in Taylor, Texas, is facing a similar situation. The company had planned to invest $44 billion to build two semiconductor plants and an advanced packaging R&D center. Under the CHIPS Act, Samsung was set to receive $6.4 billion in U.S. government subsidies.

Although construction on the first plant is underway after multiple delays, with completion targeted for 2026 and advanced sub-5nm processes in development, Digital Daily reported that the start of construction for the second plant, originally expected this year, has been postponed. As a result, related orders have also been delayed.

As reported by Digital Daily, a representative from Samsung Electronics stated that the company is unable to provide an official update on its production line operations, but it may adjust plant construction plans in response to fluctuations in demand.

(Photo credit: Samsung)

SK Hynix announced today that it has commenced mass production of the world’s first 12-layer HBM3E product with 36GB of capacity, the largest for any HBM currently available, according to the company.

SK Hynix stated that it plans to deliver these mass-produced units to customers by year-end, marking another technological milestone just six months after shipping its 8-layer HBM3E product in March.

The company also emphasized that it remains the only firm globally to have developed and supplied the entire HBM lineup, from HBM1 to HBM3E, since debuting the world’s first HBM in 2013.

The 12-layer HBM3E meets the highest global standards in speed, capacity, and stability—all critical for AI memory, SK Hynix said. The memory’s operational speed has been increased to 9.6 Gbps. When paired with a single GPU running four HBM3E units, AI models like ‘Llama 3 70B’ can process 70 billion parameters 35 times per second.

SK Hynix has boosted capacity by 50% by stacking 12 layers of 3GB DRAM chips at the same thickness as the previous 8-layer product. To achieve this, each chip was made 40% thinner and stacked using TSV technology.

By employing its advanced MR-MUF process, SK Hynix claims to have resolved structural challenges posed by stacking thinner chips. This allows for 10% better heat dissipation and enhanced stability and reliability through improved warpage control.

“SK hynix has once again broken through technological limits demonstrating our industry leadership in AI memory,” said Justin Kim, President (Head of AI Infra) at SK hynix. “We will continue our position as the No.1 global AI memory provider as we steadily prepare next-generation memory products to overcome the challenges of the AI era.”

(Photo credit: SK Hynix)

Driven by the massive demand for AI chips, advanced packaging is in short supply, and the development of three major advanced packaging technologies CoWoS, SoIC, and FOPLP is booming. In addition, two major billion-dollar projects in China have made recent progress, further advancing the region’s advanced packaging and testing industry.

According to estimates from TrendForce, revenue from 2.5D packaging services offered by foundries is expected to grow by over 120% annually by 2025. Although this segment will account for less than 5% of overall foundry revenue, its importance is steadily increasing.

HT-Tech and Tong Fu Advance Billion-Dollar Advanced Packaging Projects

On September 22, the second phase of the HT-Tech Nanjing Integrated Circuit Advanced Packaging and Testing Industrial Base project broke ground in the Pukou District of Nanjing.

The first phase of this project began production in July 2020, achieving a production value of 2.9 billion yuan in 2023. HT-Tech Group has decided to invest another 10 billion yuan to launch the second phase, aiming to build 200,000 square meters of factories and supporting facilities, introduce high-end production equipment, and create an internationally advanced integrated circuit packaging and testing production line. The products will be widely used in memory, RF, computing power, AI, and other fields.

In addition, two sub-projects of Tongfu’s advanced packaging base have also made progress. Construction of the Tongfu Tongda Advanced Packaging Base project has officially begun, and on the same day, the first piece of equipment for the second phase of Tongfu Tongke’s Memory project was installed.

The total investment in the Tongfu Tongda Advanced Packaging Base project is 7.5 billion yuan, covering 217 acres. The project is expected to be fully operational by April 2029, focusing on communications, memory, computing power, and other application areas, with a particular emphasis on multi-layer stacking, flip-chip, wafer-level, and panel-level packaging, all key products supported by national policy.

The second phase of the Tongfu Tongke Memory project adds 8,000 square meters of cleanroom space, and once operational, will provide 150,000 wafers per month. Additionally, 160 million yuan will be invested in equipment, primarily critical for the mass production of high-end products such as embedded FCCSP and uPOP, better meeting the domestic demand for high-end memory products in mobile phones, solid-state drives, servers, and more.

Progress in Advanced Packaging by TSMC, Amkor, and ASE

After purchasing the fourth factory of Innolux in Southern Taiwan Science Park (STSP) in mid-August, TSMC has swiftly begun construction, aiming to have the first phase of the new factory and equipment ready before the Chinese New Year in 2025. TSMC has integrated the new facility into its CoWoS production plan, striving to double total capacity by 2025.

TSMC expects CoWoS capacity to continue expanding rapidly through 2026, with an annual compound growth rate of over 50% from 2022 to 2026.

In terms of expanding CoWoS capacity, TSMC is also strengthening its cooperation with OSAT (Outsourced Semiconductor Assembly and Test). It is reported that TSMC has outsourced the critical CoW process in the front-end and the WoS process in the back-end to SPIL (a subsidiary of ASE Technology Holding) and Amkor.

At Amkor, the company has reportedly completed investments to expand the 2.5D advanced packaging capacity at its K5 factory in Songdo, Incheon, South Korea, tripling its capacity compared to the second quarter of last year.

At ASE, plans are underway for the Central Taiwan Science Park’s main plant and its second factory. Additionally, ASE’s wholly-owned subsidiary SPIL recently repurchased the Tanzi Factory. Furthermore, existing factories, such as the one in Zhongshan, are being reallocated. Once the cleanrooms are complete, equipment installation will begin, increasing capacity by at least 20%.
、
(Photo credit: TSMC)

While South Korean memory giants Samsung Electronics and SK hynix saw their sales in China double in the first half of this year, the country as a whole seems to heavily rely on China for essential semiconductor raw materials as well, with silicon, germanium, gallium and indium seeing the largest increase, according to a report by the Korea Eximbank Overseas Economic Research Institute on September 24 cited by Business Korea.

Despite the efforts to diversify supply chains, the report highlights the growing reliance of South Korea on China for critical semiconductor raw materials. For instance, the importance of silicon, a vital component in silicon wafer production, has been increasing, as the country’s reliance on China for the ingredient rose from 68.8% to 75.4% in 2022, the report states.

Meanwhile, South Korea’s reliance on rare earths, which are used in semiconductor abrasives, is also said to be on the rise, the report notes. The reliance on tungsten, crucial for semiconductor metal wiring, experienced a slight increase as well.

It is worth noting that since August of last year, the Chinese government has imposed export restrictions on critical minerals, including germanium and gallium, as a counteract to U.S. export sanctions. According to the U.S. Geological Survey, China produces 98% of the world’s gallium and 60% of germanium.

Even before the sanction, there is a significant rise in South Korea’s dependence on China for these critical minerals. Business Korea notes that the country’s dependence on germanium, which is frequently used in next-gen compound semiconductors, surged by 17.4 percentage points to 74.3% in 2022.

In addition, reliance on gallium and indium increased by 20.5 percentage points to 46.7%, according to the report.

Under the scenario of China’s export restrictions on key minerals, which were implemented in August and December of last year, the local production by major Chinese companies has not significantly declined, the report notes.

For instance, Samsung’ NAND flash facility in Xi’an, China, has increased its share for the company’s total NAND capacity during the past few years, from 29% in 2021 to 37% in 2023, with expectations to reach 40% this year, according to the report.

Read more

• [News] China’s Gallium and Germanium Export Restrictions Risk Chip Production Shortages
• [News] Chip War Escalates: China to Enforce State Ownership of All Rare Earth Materials in October

(Photo credit: Samsung)