As AI giant NVIDIA is said to delay its upcoming Blackwell series chips for months, which are now expected to hit the market around early 2025, the related semiconductor supply chain is experiencing a reshuffle. According to a report by the Korea Economic Daily, Samsung Electronics, which is eager to expand its market share for HBM3 and HBM3e, is likely to emerge as a major beneficiary in addition to AMD.

In March, NVIDIA introduced the Blackwell series, claiming it could enable customers to build and operate real-time generative AI on trillion-parameter large language models at up to 25 times less cost and energy consumption compared to its predecessor.

However, according to The Information, last week, NVIDIA informed major customers, including Google and Microsoft, that the shipments of its Blackwell AI accelerator would be delayed by at least three months due to design flaws.

Blackwell Delayed Potentially due to Design Flaws and TSMC’s Capacity Constraints

Tech media Overclocking points out that the defect is related to the part connecting the two GPUs, and creates problems for NVIDIA’s dual GPU versions, including the B200 and the GB200.

The delay has prompted tech companies to look for alternatives from NVIDIA’s competitors, such as AMD, according to the Korea Economic Daily. Microsoft and Google have already been working on next-generation products with AMD. For instance, Microsoft has purchased the MI300X, an AI accelerator from the US fabless semiconductor designer, the report says.

Samsung to Benefit thanks to the Collaboration with AMD

Samsung, as its HBM3 received AMD MI300 series certification in 1Q24, and is likely to provide HBM3e chips to AMD afterwards, is expected to benefit. Citing a semiconductor industry source, the Korea Economic Daily notes that as it is very risky for a single company to dominate the AI chip supply chain, the situation will create opportunities for Samsung and AMD.

It is also worth noting that Samsung’s HBM3 has passed NVIDIA’s qualification earlier, and would be used in the AI giant’s H20, which has been developed for the Chinese market in compliance with U.S. export controls.

According to TrendForce’s forecast in mid-July, the shipment share of AI servers equipped with self-developed ASIC chips in 2024 is expected to exceed 25%, while NVIDIA owning the lion’s share of 63.6%. AMD’s market share, on the other hand, is projected to reach 8.1% in 2024.

Read more

• [News] Foxconn Remains Unfazed Despite NVIDIA’s GB200 Delay, with Customers Aim to Secure Their Orders
• [News] Samsung’s 8-layer HBM3e to Start Mass Production in Q3, Driving HBM Sales to Soar 3-5 Times in 2H24

(Photo credit: NVIDIA)

Semiconductor giants like Intel, TSMC, Samsung and Micron have received huge amount of grants from the U.S. government, funded through the CHIPS and Science Act. However, chip making equipment maker Applied Materials is said to be in a different scenario. According to reports by Bloomberg and Tom’s Hardware, the company’s application to gain U.S. funding for a USD 4 billion R&D center in Silicon Valley was rejected by U.S. Department of Commerce.

The reports note that Applied Materials had announced plans to build the facility a year ago, as it tried to seek government subsidies through the CHIPS and Science Act. The facility was scheduled for completion in 2026.

However, according to sources familiar with the matter, Commerce Department officials turned down the plan on Monday, stating that project did not meet the eligibility criteria, Bloomberg reports. This decision marks a major setback for the company’s efforts to establish a significant facility in Silicon Valley, which it aims to develop next-generation chip making tools.

In addition, though it is reported that as there are over 670 companies with interests in the gaining the fund under the CHIPS and Science Act, and the Commerce Department has warned that limited resources will force it to reject many applications, the rejection of Applied Materials’ project is particularly unexpected. For it is a U.S. semiconductor company, and the project closely aligns with the Biden administration’s goals of revitalizing the domestic semiconductor industry.

It is worth noting that though the U.S. keeps tightening the export controls on the semiconductor sector, major chip equipment makers seem to become increasingly dependent on the Chinese market. From February to April, China accounted for 43% of the total sales of Applied Materials, a 22 percentage point increase YoY.

Applied Materials has reportedly received subpoenas from the US Securities and Exchange Commission as well as the US Attorney’s Office of the District of Massachusetts in February, and said to be under investigation for allegedly sending equipment to SMIC, China’s leading chip maker, through South Korea without export licenses.

The CHIPS and Science Act, signed into law in August 2022, allocated approximately USD 280 billion in new funding to enhance domestic chip making research and development.

Previously, the U.S. government announced that Intel would receive USD 8.5 billion in federal subsidies and USD 11 billion in loans. On the other hand, US administration is set to provide USD 6.6 billion and USD 6.4 billion in aid to TSMC and Samsung, respectively.

Read more

• [News] US Reportedly Expands Investigation into Applied Materials, Possibly due to Concerns of Shipments to SMIC
• [News] China Contributes Over 40% of U.S. Chip Equipment Makers’ Revenue despite Export Controls

(Photo credit: Applied Materials)

As the battle of HBM intensifies between memory giants, the competition of NAND is also heating up. According to a report by Korean media outlet etnews, SK hynix is developing 400-layer NAND flash memory, aiming to get the technology ready for mass production by the end of 2025.

Citing sources familiar with the matter, the report notes that SK hynix is currently working with supply chain partners to develop process technologies and equipment needed for 400-layer and above NANDs. As the company plans to apply hybrid bonding to achieve the breakthrough, many packaging materials and components suppliers are expected to enter the new supply chain.

According to the report, SK hynix is reviewing new materials for bonding and various technologies for connecting different wafers, including polishing, etching, deposition, and wiring. With the goal of getting the technology and infrastructure ready by the end of next year, full-scale production for the 400-layer NAND is anticipated to begin in the first half of 2026.

Currently, the Big Three in the memory sector are all trying to push the boundaries on the layers of NAND. Earlier in April, Samsung confirmed that it has begun mass production for its one-terabit (Tb) triple-level cell (TLC) 9th-generation vertical NAND (V-NAND), with the number of layers reaching 290. For now, the company aims to stack V-NAND to over 1000 layers by 2030.

Micron, on the other hand, has announced the 2650 client SSD, its first product built from 276-layer 3D NAND on July 30^th. Japanese memory chipmaker Kioxia, after successfully increasing the number of 3D NAND layers to 218 in 2023, even stated that achieving a 1,000-layer level by 2027 would be possible.

In August, 2023, SK hynix showcased its sample of the world’s first 321-layer NAND product. Now, as the limit is expected to be pushed up to 400 layers, the company plans to apply hybrid bonding to the manufacturing, which adopts a “wafer-to-wafer” (W2W) structure, etnews notes.

According to the report, until now, SK hynix has been stacking cells on top of the peripherals, the driving circuit area, using the method of “Peripheral Under Cell (PUC)” to manufacture NAND. The structure is similar to a mixed-use high-rise apartment where the peripheral (commercial space) is at the bottom and the cells (residential units) are stacked on top.

However, as the number of NAND layers increases, the peripheral is prone to be damaged during the cell stacking process due to the high heat and pressure generated during the cell process, the report explains.

Therefore, SK hynix plans to apply hybrid bonding to overcome the issues. By implementing cells and peripherals on separate wafers and then bonding the two wafers together, the method allows the peripheral wafer that drive the cells to be separately manufactured, thus enabling a stable increase in NAND layers.

Regarding the progress on the development of 400-layer NAND, SK hynix stated that it cannot confirm details about its technology development or mass production timeline, the report notes.

Read more

• [News] SK Hynix Plans to Manufacture 3D NAND over 400 Layers at -70°C with the Help of TEL
• [News] Outpacing Samsung or the End of Race? Kioxia Aims 1000-layer NAND by 2027

(Photo credit: SK hynix)

Samsung Electronics, which has been surround by concerns that its HBM3e products are still struggling to pass NVIDIA’s qualifications, has confirmed in its second quarter earnings call that the company’s fifth-generation 8-layer HBM3e is currently undergoing customer valuation, and is scheduled to enter mass production in the third quarter, according to a report by Business Korea.

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.

Driven by the momentum, the report from Business Korea, citing an official speaking at the conference call on July 31^st, states that the share of HBM3e chips within Samsung’s HBMs is anticipated to surpass the mid-10 percent range in the third quarter. Moreover, it is projected to speedily grow to 60% by the fourth quarter.

According to Samsung, its HBM sales in the second quarter already grew by around 50% from the previous quarter. Being ambitious about its HBM3 and HBM3e sales, Samsung projects its HBM sales will increase three to five times in the second half of 2024, driven by a steep rise of about two times each quarter.

Samsung has already taken a big leap on HBM as its HBM3 chips are said to have been cleared by NVIDIA last week. According to a previous report by Reuters, Samsung’s HBM3 will initially be used exclusively in the AI giant’s H20, which is tailored for the Chinese market.

On the other hand, the South Korean memory giant notes that it has completed the preparations for volume production of its 12-layer HBM3e chips. The company plans to expand the supply in the second half of 2024 to meet the schedules requested by multiple customers, according to Business Korea. The progress of its sixth-generation HBM4 is also on track, scheduled to begin shipping in the second half of 2025, Business Korea notes.

Samsung Electronics reported higher-than-expected financial results in the second quarter, with a six-fold year-on-year increase in net income, soaring from KRW 1.55 trillion won (USD 1.12 billion) to KRW 9.64 trillion (USD 6.96 billion), as demand for its advanced memory chips that are crucial for AI training remained strong.

SK hynix, as the current HBM market leader, has expressed its optimism in securing the throne as well. The company reportedly expects its HBM3e shipments to surpass those of HBM3 in the third quarter, with HBM3e accounting for more than half of the total HBM shipments in 2024. In addition, it expects to begin supplying 12-layer HBM3e products to customers in the fourth quarter.

Micron, on the other hand, has reportedly started mass production of 8-layer HBM3e as early as in February. The company reportedly plans to complete preparations for mass production of 12-layer HBM3e in the second half and supply it to major customers like NVIDIA in 2025.

Read more

• [News] Samsung’s Q2 Profits Soar to USD 7.5 Billion, Seeing Strong Demand for HBM, DDR5 and Server SSD in 2H24
• [News] Samsung Sees Solid Demand for 3nm in 2H24, Aiming a Ninefold Increase in AI/HPC Sales by 2028

(Photo credit: Samsung)

In addition to the strong memory momentum which contributes to Samsung’s soaring profits in the second quarter, the tech giant’s progress on the foundry and chip business also attracts attention. According to its press release on July 31st, Samsung expects its foundry revenue growth to outpace the market in 2024 on the back of the full-scale mass production of second-generation 3nm GAA technology.

Earlier in May, Samsung announced the tape-out of its first 3nm mobile SoC, which used the Synopsys.ai EDA suite to verify the design. It signifies a major milestone as it is the first smartphone AP taped out with its 3nm GAA process. Before that, Samsung’s SF3E node has only been utilized for cryptocurrency mining chips.

The Korean semiconductor heavyweight now notes that the initial response to the new SoC for wearables, which features the industry’s first 3nm technology, has been favorable, and adoption of 3nm SoCs by key customers is expected to expand in the second half of the year. It also plans to ensure a stable supply of the Exynos 2500 for flagship models.

It is worth noting that another focus for Samsung in 2H24 will be the expansion for the application of 200-megapixel sensors from main wide camera to tele cameras. Samsung states it plans to expand sales of DDI products with the start of mass production of new models for a customer based in the US. According to an earlier report by The Verge, Apple may begin using Samsung camera sensors as early as 2026, ending Sony’s decade-long role as the exclusive supplier of the phone’s camera sensors.

Samsung also draws an ambitious roadmap, saying that it will expand its order intake for AI and HPC applications, targeting a fourfold increase in the customer base and a ninefold increase in sales by 2028 from the levels in 2023.

Samsung announced its financial results for the second quarter today, posting KRW 74.07 trillion in consolidated revenue and operating profit of KRW 10.44 trillion (approximately USD 7.5 billion). Its DS Division posted KRW 28.56 trillion in consolidated revenue and KRW 6.45 trillion in operating profit for the second quarter, posting a 94% and 1081% YoY growth, respectively.

Its Foundry Business saw improved earnings as a result of increased demand across applications. Due to higher orders for sub-5nm technology, the number of AI and HPC customers increased twofold from a year earlier. The Foundry Business also distributed the process development kit (PDK) for 2nm Gate-All-Around (GAA) technology to customers ahead of mass production in 2025.

On July 9^th, Samsung confirmed that it has received the first client for its 2nm process, and will provide turnkey semiconductor solutions using the 2nm process and the advanced 2.5D packaging technology Interposer-Cube S (I-Cube S) to Japanese AI company Preferred Networks.

Read more

• [News] Samsung’s Q2 Profits Soar to USD 7.5 Billion, Seeing Strong Demand for HBM, DDR5 and Server SSD in 2H24
• [News] Samsung Officially Secures Its First 2nm Order, Making AI Chips for Preferred Networks

(Photo credit: Samsung)