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)

In 2024, the storage market is experiencing dynamic changes, with many positive developments, including rising contract prices, significant revenue growth for manufacturers, and multiple breakthroughs in technology. Amid this, major storage companies are gearing up for new challenges, especially as the NAND flash memory sector faces an impending shift.

How Far Have Major Companies Progressed with NAND Flash Technology?

This year, major storage manufacturers like Samsung, Micron, and SK Hynix have all made notable advancements in NAND flash technology.

In terms of NAND cell technology, Samsung has become the first in the industry to mass-produce its 9th generation V-NAND with QLC technology. On September 12, Samsung announced it had begun mass production of its 1Tb QLC (Quad-Level Cell) 9th generation vertical NAND (V-NAND), incorporating several groundbreaking technologies.

From a technological innovation perspective, Samsung’s 9th generation QLC V-NAND employs its unique channel hole etching technology to achieve the industry’s highest stack height with a dual-stack structure. Leveraging the expertise of TLC 9th generation V-NAND, the cell area and peripheral circuits are optimized, resulting in an 86% higher bit density than the previous generation.

Compared to earlier versions, the design of Samsung’s 9th generation QLC V-NAND improves data retention performance by about 20%, enhancing product reliability. Writing performance has doubled, and data input/output speeds have increased by 60%. Additionally, its low-power design reduces power consumption for both read and write operations by approximately 30% and 50%, respectively. This is achieved by sensing only the necessary bit lines (BL) to minimize power usage.

In terms of applications, Samsung plans to expand the use of the 9th generation QLC V-NAND from branded consumer products to mobile universal flash storage (UFS), PC, and server SSDs to meet the demands of customers, including cloud service providers.

Sung Hoi Hur, Executive Vice President and Head of Flash Product and Technology at Samsung Electronics, stated that as the enterprise SSD market grows rapidly and demand for AI applications increases, the company will continue to strengthen our leadership in the high-capacity, high-performance NAND flash market through 9th-generation QLC and TLC V-NAND.

However, at present, the mainstream products in the market are still TLC NAND flash memory particles. On August 6, SK Hynix’s Solidigm launched PCIe 5.0 data center SSDs, the D7-PS1010/1030 series, based on SK Hynix’s 176-layer 3D TLC NAND.

On September 11, SK Hynix announced the development of its high-performance SSD “PEB110 E1.S” for data centers, available in 2TB, 4TB, and 8TB versions. Currently undergoing validation with global data center customers, SK Hynix plans to begin mass production in the second quarter of next year.

On the other hand, Micron announced in late July that its SSD products featuring 9th generation (G9) TLC NAND technology had entered mass production, targeting personal devices, edge servers, enterprises, and cloud data centers. Micron’s G9 NAND achieves a data transfer rate 50% faster than current NAND technology used in SSDs. Its per-chip write and read bandwidths are 99% and 88% higher, respectively, than other NAND solutions. The Micron 2650 NVMe SSD, based on G9 NAND, achieves near-PCIe 4.0 performance levels, with a sequential read speed of up to 7,000 MB/s.

Micron also launched its new data center SSD, the 9550 NVMe SSD, featuring 232-layer 3D TLC NAND. It supports various AI workloads, offering a sequential read speed of 14.0 GB/s and a write speed of 10.0 GB/s—67% higher than competitive SSDs. The 9550 SSD’s random read speed reaches 3,300K IOPS, 35% higher than competitors, with random write speeds 33% higher.

The future of QLC SSDs is promising

Industry information indicates that NAND Flash, the core medium for data storage, is vital for SSD performance. Current SSDs use both TLC (Triple-Level Cell) and QLC flash.

In the AI era, there is a growing demand for storage, with SSDs playing a critical role. According to TrendForce, SSDs not only store model parameters during AI model training but also create checkpoints to save progress, making them crucial for high-speed data transfer and durability. As a result, customers primarily opt for 4TB/8TB TLC SSDs to meet the rigorous demands of AI training processes.

QLC SSDs, however, are gaining attention due to their higher storage density, which optimizes server space and reduces energy consumption. They can help large-scale data centers lower their total cost of ownership (TCO) while still meeting high-performance storage needs. Industry experts predict that as more data is generated in the form of videos and images, requiring larger storage capacities, TLC/QLC SSDs of 16TB or more will become the primary products for AI inference applications.

According to TrendForce, AI-related SSD procurement is expected to exceed 45 exabytes (EB) in 2024, with SSD demand in AI servers projected to grow by over 60% annually in the coming years. The share of AI SSDs within the NAND Flash market could rise from 5% in 2024 to 9% in 2025.

The Changing Landscape of the NAND Flash Market

On September 9, TrendForce’s latest research  indicates that in the second quarter of 2024, Samsung maintained its global leadership in the NAND Flash market with a 36.9% market share, up 0.2% from the previous quarter. SK Group followed with a 22.1% share, down 0.1%. Other key players include Kioxia (13.8%), Micron (11.8%), and Western Digital (10.5%).

In terms of revenue, Samsung, SK Group, Kioxia, Micron, and Western Digital all experienced quarter-on-quarter growth in NAND Flash revenues during the second quarter. Overall, NAND Flash revenue increased by 14% in the second quarter.

TrendForce indicates that as the inventory adjustments for server endpoints near completion and AI drives demand for high-capacity storage products, NAND Flash prices continued to rise in Q2 2024. However, due to high inventory levels at PC and smartphone manufacturers, NAND Flash bit shipments decreased by 1% quarter-over-quarter. Despite this, the average selling price increased by 15%, with total revenue reaching USD 16.796 billion, a 14.2% increase from the previous quarter.

Looking ahead to Q3, TrendForce expects that all NAND Flash suppliers have returned to profitability as of Q2 and plan to expand production capacity in Q3 to meet strong demand from AI and servers. However, due to weak market performance in the PC and smartphone sectors in the first half of the year, it is challenging to boost NAND Flash shipments. It is estimated that the average selling price of NAND Flash products will increase by 5% to 10% in Q3, while bit shipments may decrease by at least 5% due to a lack of peak season demand. Industry revenue is expected to remain roughly the same as the previous quarter.

(Photo credit: Samsung)

In recent years, the U.S., Japan and the Netherlands, have increasingly expanding restrictions on China in semiconductor technology. South Korea, on the other hand, has been cautiously responding to U.S. demands due to its significant dependence on the Chinese market.

Yet, according to a report by South Korean media outlet The Korea Herald, the U.S. is increasing pressure on South Korea to comply with its export controls to China.

At the Korea-U.S. Economic Security Conference 2024 held in Washington, D.C. on September 10th, U.S. Commerce Department Undersecretary Alan Estevez called on South Korea’s two leading HBM manufacturers, Samsung and SK hynix, to align with U.S. export controls on China. He urged that their production capacity be reserved for supplying advanced chips to allied nations, rather than competitors such as China.

Estevez emphasized his appreciation for South Korea’s long-standing cooperation with the U.S., but pointed out that since AI can be used for military purposes, it is crucial to prevent China from acquiring advanced chips to train AI models.

South Korea’s Trade Minister Cheong In-kyo responded that while they will discuss the matter with the U.S., export controls have a significant impact on South Korea’s businesses and economy.

Some industry sources cited by The Korea Herald have further pointed out that the direct export volume of chips from Samsung and SK hynix to China is not significant, so the actual impact may be limited.

However, per a previous Reuters report cited sources, it’s indicated that about 30% of Samsung’s HBM chip sales in the first half of this year were to China.

The Korea Institute for Industrial Economics and Trade noted that, unlike Japan and the Netherlands, South Korea cannot fully align with U.S. export control measures due to its significant reliance on exports to China.

Per the Chosun Daily citing data from South Korea’s Ministry of Trade, Industry, and Energy and the Korea International Trade Association, it’s shown that in July of this year, South Korea’s exports to China increased by 14.9% year-on-year to USD 11.4 billion, the highest since October 2022. Notably, memory exports surged 89% year-on-year to USD 6.8 billion.

Semiconductor exports saw particularly strong growth, with chip exports rising 49% year-on-year. In June this year, Korea’s memory exports also amounted to USD 8.8 billion, accounting for 65.8% of total semiconductor exports, which reportedly represents the highest proportion in two years since December 2021.

These figures reflect South Korea’s robust performance in the chip sector and the strong demand from the Chinese market for Korean semiconductors and other ICT products.

Meanwhile, due to the U.S.’s strict restrictions on chip manufacturing technology, China is striving for breakthroughs in the HBM field.

The HBM market is currently dominated by South Korea’s SK hynix, Samsung Electronics, and the U.S.’s Micron, all of which are producing the latest standard HBM3 chips.

However, a report from Tom’s Hardware, citing industry sources, has indicated that Chinese companies, including CXMT, have made progress in developing HBM and are in the early stages of production. Huawei is also collaborating with other Chinese companies, with plans to produce HBM2 chips by 2026.

Read more

• [News] ChangXin Memory Technologies in China Has Reportedly Begun Mass Production of HBM2
• [News] Progress Made in Chinese Consortium’s Development of HBM, Reportedly Targeting HBM2 Production by 2026

(Photo credit: Samsung)

According to sources cited by Indian media outlet Financial Express, Apple is said to be in talks with Micron, Tata Group, and other Indian chip manufacturers to procure USD 12 billion worth of chips locally for iPhones produced in India.

Reportedly, Apple plans to shift 26% of its iPhone production to India by 2026. Industry sources cited by the report further suggest that by then, Apple will become the largest single buyer of Indian-made semiconductors, surpassing any other sectors such as defense, aerospace, and automotive.

Sources further point out that if Micron and Tata are able to produce chips that meet Apple’s requirements, a significant portion of the chips needed for iPhones will come from these companies, potentially creating substantial opportunities for the Indian semiconductor industry.

Before the outbreak of the pandemic, Apple iPhones and almost all other consumer products were manufactured in China. Following the introduction of the Production-Linked Incentive (PLI) scheme by the Indian government, potentially prompting Apple to begin shifting its production lines.

In 2022, the Indian government launched a USD 10 billion PLI scheme to stimulate domestic semiconductor production. To date, India has approved five chip manufacturing projects with a total value of roughly USD 18 billion and has reserved USD 1.2 billion for future projects.

Meanwhile, Micron’s chip plant in Gujarat, India, is expected to begin operations this year, marking the company’s first chip facility in India.

Additionally, Tata Group, in collaboration with PSMC, plans to build a packaging plant in Gujarat, with operations slated to commence in 2026, producing chips using 28nm, 40nm, 55nm, 90nm, and 110nm nodes.

Read more

• [News] 300mm Wafer Fab Construction Picks up Steam Worldwide
• [News] Reducing Reliance on China, Apple to Produce High-End iPhones in India for the First Time

(Photo credit: Tata Group)

After its 8-Hi HBM3e entered mass production in February, Micron officially introduced the 12-Hi HBM3e memory stacks on Monday, which features a 36 GB capacity, according to a report by Tom’s Hardware. The new products are designed for cutting-edge processors used in AI and high-performance computing (HPC) workloads, including NVIDIA’s H200 and B100/B200 GPUs.

It is worth noting that the achievement has made the US memory chip giant almost on the same page with the current HBM leader, SK hynix. Citing Justin Kim, president and head of the company’s AI Infra division at SEMICON Taiwan last week, another report by Reuters notes that SK hynix is set to begin mass production of its 12-Hi HBM3e chips by the end of this month.

Samsung, on the other hand, is said to have completed NVIDIA’s quality test for the shipment of 8-Hi HBM3e memory, while the company is still working on the verification of its 12-Hi HBM3e.

Micron’s 12-Hi HBM3e memory stacks, according to Tom’s Hardware, feature a 36GB capacity, a 50% increase over the previous 8-Hi models, which had 24GB. This expanded capacity enables data centers to handle larger AI models, such as Meta AI’s Llama 2, with up to 70 billion parameters on a single processor. In addition, this capability reduces the need for frequent CPU offloading and minimizes communication delays between GPUs, resulting in faster data processing.

According to Tom’s Hardware, in terms of performance, Micron’s 12-Hi HBM3e stacks deliver over 1.2 TB/s. Despite offering 50% more memory capacity than competing products, Micron’s HBM3e consumes less power than the 8-Hi HBM3e stacks.

Regarding the future roadmap of HBM, Micron is said to be working on its next-generation memory solutions, including HBM4 and HBM4e. These upcoming memory technologies are set to further enhance performance, solidifying Micron’s position as a leader in addressing the increasing demand for advanced memory in AI processors, such as NVIDIA’s GPUs built on the Blackwell and Rubin architectures, the report states.

Read more

• [News] SK hynix to Begin 12-Stack HBM3e Mass Production, Marking Key Moment in HBM Battlefield
• [News] Samsung Likely Emerges as the Pacemaker for the AI Market if It Secures HBM3e Supply to NVIDIA

(Photo credit: Micron)