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)

On August 13th, as per a report from Wallstreetcn citing industry sources, it’s indicated that SK hynix has raised the price of its DDR5 DRAM by 15% to 20%. Per the sources, the price hike by hynix is primarily due to the production capacity being squeezed by HBM3/3e. Additionally, the increased orders for AI servers downstream have also strengthened SK hynix’s resolve to raise DDR5 prices.

According to industry sources cited by Economic Daily News, for Taiwanese manufacturers, Nanya Technology has recently started mass production of DDR5, just in time to benefit from this price surge. Module makers such as ADATA and Team Group are also likely to see gains from low-cost inventory.

Nanya Technology has begun shipping its 16Gb DDR5, developed using its 1B process. Nanya Technology is optimistic that the DRAM market is on a clear path to recovery. This may due to last year’s production cuts by the three major memory manufacturers—Samsung, SK hynix, and Micron—as well as the strong demand for HBM driven by generative AI. The resulting chain reaction is expected to positively impact various types of DRAM.

SK hynix previously announced that its entire HBM production capacity for 2024 has been fully booked, with almost all of its 2025 capacity also sold out. To meet customer demand, SK hynix plans to convert over 20% of its existing DRAM production lines to mass-produce HBM.

Samsung, on the other hand, is said to be actively trying to catch up with SK hynix, looking to allocate around 30% of its DRAM production capacity to HBM.

The significant adjustments by Samsung and SK hynix to their production lines have severely squeezed the capacity for DDR4 and DDR5 DRAM, potentially leading to a sharp reduction in supply and causing prices to rise. Reportedly, SK hynix’s price increase for DDR5 primarily targets contract prices.

Read more

• [News] Nanya Technology Chairman Names 10nm 1B Process as Key Expansion Focus for the Year
• [News] DRAM Price Surge Halts, Could HBM Demand Drive Price Increases?

(Photo credit: Nanya Technology)

In the memory sector, the era beyond 10nm seems to be around the corner, as giants have revealed plans on the production of 1c DRAM. SK hynix, according to a report by The Elec, is said to develop a 4F2 (square) DRAM in order to reduce the high cost associated with the extreme ultraviolet (EUV) processes since the commercialization of 1c DRAM.

During an industry conference on Monday, Seo Jae Wook, a SK hynix researcher, questioned the profitability of producing DRAM with EUV technology. Instead, he said that SK hynix is exploring the possibility of manufacturing future DRAMs using vertical gate (VG) or 3D DRAM structure, according to the report.

Sources cited by the report indicate that Samsung and SK hynix are both targeting to apply 4F2 for DRAM at the 10nm node and beyond. According to Seo, by applying VG or the 3D DRAM structure, the cost of the EUV processes can be reduced by half.

SK hynix’s VG, or the so-called 4F2 by memory makers internally, is similar to the vertical channel transistor (VCT) of its South Korean competitor Samsung, The Elec explains.

In short, unlike the traditional 2D DRAM, which only uses the horizontal plane, 4F2, or 3D DRAM, is a cell array structure where transistors are stacked vertically. In the structure, various parts including the source, gate, drain, and capacitor are stacked from bottom to top. As the word line connects to the gate, the bit line connects to the source.

According to the report, arranging the cell array in this manner can reduce the die surface area by 30% compared to 6F2 DRAM.

However, according to a previous report by TweakTown, though the 4F2 design possesses the advantages of being compact and power-efficient, its complexity demands extreme precision in fabrication, higher-quality materials for production, and extensive research to make it scalable and suitable for mass production.

According to TweakTown, Samsung is anticipated to finalize the initial development of VCT DRAM by 2025, with 3D DRAM expected to enter the market by 2030.

Read more

• [News] SK hynix Reportedly Develops NAND over 400 Layers, Eyeing Mass Production Ready by 2025
• [News] SK hynix CEO: Demand for Memory Chips to Remain Robust until 1H25, Driven by HBM

(Photo credit: SK hynix)

Facing increased market demand and the ongoing recovery of the memory industry, a report from Korean media outlet ETNews has reported that Samsung has confirmed its investment plan for the 6th-Generation DRAM production line at the Pyeongtaek P4 plant, with the goal of starting mass production in June 2025.

Reportedly, the 6th-generation DRAM, known as ‘1c ,’ is a next-generation DRAM utilizing 10nm-class technology. Despite it is said to be a product that has not yet been commercialized in the global semiconductor industry, both Samsung Electronics and SK hynix are already preparing for mass production.

Samsung’s Pyeongtaek P4 is a comprehensive semiconductor production center, divided into four phases.

As per ETNews, Samsung plans to initiate 1c DRAM production by the end of this year. The company is said to be considering launching HBM4 using 1c DRAM in the second half of 2025.

Given that HBM consumes significantly more DRAM than traditional memory, it is speculated by the report that Samsung’s construction of the 1c DRAM production line at the Pyeongtaek P4 plant may also be in preparation for HBM4.

As per TrendForce’s latest report on the memory industry, it’s revealed that DRAM and NAND Flash revenues are expected to see significant increases of 75% and 77%, respectively, in 2024, driven by increased bit demand, an improved supply-demand structure, and the rise of high-value products like HBM.

Furthermore, TrendForce also reports that Samsung’s P4L facility will be the key site for expanding memory capacity starting in 2025, starting with NAND production. Equipment installation for DRAM is expected to begin in mid-2025, with mass production of 1c nanometer DRAM slated to commence in 2026.

Read more

• [News] Samsung Develops Custom HBM with Tech Giants, with Commercialization Expected in HBM4
• [News] Overview of Expansion Plans by HBM Giants

(Photo credit: Samsung)

According to a report from Korea media outlet Yonhap News Agency, South Korea’s memory export to Taiwan has surged by 225% in the first half of the year.

The primary driver of this increase is reportedly due to South Korean chipmaker SK hynix’s supply of HBM to U.S. AI chip giant NVIDIA, which packages its AI accelerators at Taiwan’s TSMC.

A researcher at the Korea Institutes for Industrial Economics & Trade, Kim Yang-paeng, also noted that the sharp increase in exports is likely related to SK hynix’s supplies for TSMC’s final packaging of AI accelerators.

The report from Economic Daily News further highlights the strong momentum in NVIDIA’s AI chip shipments, with TSMC, as the key manufacturing partner, receiving steady advanced process orders.

The report from Yonhap News Agency also cited data from the industry ministry and the Korea International Trade Association released on August 11th, showing that South Korea’s memory exports to Taiwan in the first half of the year grew by 225.7% year-on-year, reaching USD 4.26 billion.

This growth significantly outpaces the overall increase in South Korea’s memory exports, which was 88.7%. Additionally, Taiwan has become South Korea’s third-largest market for memory exports in the first half of the year, climbing two spots to surpass Vietnam and the United States.

Another Korean media outlet, The Korea Herald, noted that since the 2010s, South Korea’s annual memory exports to Taiwan have ranged between USD 1 billion and 4 billion. The latest data indicates that this year’s export volume may set a new record, potentially reaching USD 8 billion.

Read more

• [News] SK hynix CEO: Demand for Memory Chips to Remain Robust until 1H25, Driven by HBM
• [News] Samsung’s 8-layer HBM3e to Start Mass Production in Q3, Driving HBM Sales to Soar 3-5 Times in 2H24

(Photo credit: SK hynix)