BTC $62,700.00 -1.81%
ETH $1,869.23 -1.26%
BNB $604.46 -0.87%
XRP $1.00 -0.45%
SOL $75.28 -1.51%
TRX $0.3326 -0.52%
DOGE $0.0695 -0.80%
ADA $0.1785 -3.35%
BCH $200.36 -6.16%
LINK $8.77 -0.49%
HYPE $55.59 -4.50%
AAVE $85.73 -3.48%
SUI $0.6758 -2.42%
XLM $0.1598 -0.63%
ZEC $483.64 -2.27%
BTC $62,700.00 -1.81%
ETH $1,869.23 -1.26%
BNB $604.46 -0.87%
XRP $1.00 -0.45%
SOL $75.28 -1.51%
TRX $0.3326 -0.52%
DOGE $0.0695 -0.80%
ADA $0.1785 -3.35%
BCH $200.36 -6.16%
LINK $8.77 -0.49%
HYPE $55.59 -4.50%
AAVE $85.73 -3.48%
SUI $0.6758 -2.42%
XLM $0.1598 -0.63%
ZEC $483.64 -2.27%

product

All
Article
Flash

hot_img Samsung Electronics plans to repurpose the NRD-K R&D production line for wafer foundry use, aiming for mass production of 2nm base chips for HBM

According to South Korean media ZDNet Korea, Samsung Electronics is adjusting the original purpose of its NRD-K Line 2 at the Giheung campus from a research and development facility to a wafer foundry line, with the main target product being the 2nm base chips required for NVIDIA's HBM. This production line is expected to be operational in the second half of 2028, adopting a "Send-Fab" model, which involves receiving wafers from other mass production lines and performing specific processes.NRD-K is the most advanced composite research and development campus built by Samsung Electronics to seize future semiconductor technology, with a total investment of approximately 20 trillion won, planning for three production lines, of which the first line is scheduled to be completed by the end of 2024. This adjustment aims to proactively respond to the expansion of HBM demand driven by investments in AI infrastructure. Samsung plans to first apply the 2nm process to custom HBM and HBM5 to maintain performance advantages.Industry insiders point out that the scale of NRD-K Line 2 is relatively small, making it suitable for operation in the Send-Fab format. However, there are still about two years until the production line is operational, and the final purpose may be adjusted again based on market conditions. Samsung had previously considered using this production line for DRAM production, but after modifications to the relevant plans this month, DRAM capacity will be more concentrated in the Pyeongtaek campus. Equipment procurement orders have not yet been officially issued.

hot_img NVIDIA's Feynman platform is expected to go into mass production in 2028, with TSMC's A16 and SoIC expanding production simultaneously

According to DIGITIMES, NVIDIA's next-generation AI platform Feynman aims for mass production in the second half of 2028, utilizing TSMC's A16 process and incorporating SoIC 3D stacking and CPO technology. The Rubin generation focuses on integrating multiple small chips with HBM, while Feynman further develops towards 3D small chips, SoIC, and CPO. The report indicates that TSMC is accelerating the construction of related production capacity, with the original plan for SoIC monthly capacity to reach 20,000 pieces by the end of 2026, now updated to a target of 50,000 pieces by the end of 2027.The construction progress of TSMC's Chiayi AP7 and Tainan AP8 is also accelerating, with AP7 planned in 8 phases. P1 and P2 have entered installation, P3 and P4 have obtained construction permits, and P5 to P8 are still in planning. In addition to mass-producing Apple-specific WMCM, the factory will configure SoIC, CoPoS, and CPO production lines according to customer demand. NVIDIA has rapidly shifted its R&D and supply chain resources to the Feynman platform, and the spillover effect of TSMC's advanced packaging orders continues to expand, with Siliconware becoming the main testing and packaging factory for NVIDIA's CoWoS and CPO orders. The NVLink interconnect bandwidth of NVIDIA's Feynman platform is expected to exceed 1 PB/s, further improving from Rubin Ultra's 520 TB/s. TSMC's COUPE technology integrates EIC and PIC in 3D through SoIC to form a light engine, advancing optoelectronic conversion from traditional circuit boards to the internal packaging structure. Since 2026, NVIDIA has invested over $40 billion in building the AI ecosystem, covering areas such as models, wafer manufacturing, data centers, and optical communications.

hot_img Xiaohongshu has created an "AI Shopping Guide" feature that directly pushes product cards and order links in conversation scenarios

According to an exclusive report from "Du Jia," Xiaohongshu is developing a brand new "AI Shopping Guide" feature, which relies on a conversational interaction model to directly push product cards and attach order jump links in Q&A scenarios. This project is led by Daoxuan (Pan Boyuan). As of the time of publication, Xiaohongshu has not responded.Previously, Xiaohongshu launched an AI summary feature in store reputation, and this AI shopping guide is a further extension of its intelligent transaction chain. According to public data, Xiaohongshu has approximately 400 million global MAU, with 39 million users engaging in explicit purchasing behavior daily, resulting in 140 million active purchase requests, over 47 million users entering merchant live streams, and more than 170,000 merchant group chats remaining active. During the 618 period, the GMV of store broadcasts increased by 265% year-on-year, and note content drove transaction growth of over 210%.Xiaohongshu currently adopts a dual transaction model of "internal marketplace closed loop plus external traffic diversion," and there is still room for improvement in the direct order conversion rate within the platform. The launch of the AI shopping guide feature is expected to further enhance its internal e-commerce GMV and revenue.

hot_img SK Hynix will expand its NAND production capacity in Dalian by about 50%, with the second factory in Dalian expected to start production in the first half of next year

SK Hynix has restarted the construction of its second NAND flash memory factory in Dalian, expecting to increase the monthly production capacity in the region by about 50%. The new production line is designed for a monthly capacity of approximately 50,000 wafers, combined with the existing Dalian Plant 1's monthly capacity of 100,000 wafers, bringing SK Hynix's total NAND production capacity in China to about 150,000 wafers per month. The factory began construction four years ago but was paused due to the downturn in the memory chip market and U.S. export restrictions on equipment to China, having only completed the structural framework.The investment in Dalian Plant 2 is driven by SK Hynix's NAND subsidiary Solidigm, with equipment installation expected to start as early as November this year, and plans to establish a mass production system in the first half of next year. Driven by the expansion of AI data centers, demand for enterprise-level SSDs has surged, with NAND prices rising nearly tenfold compared to a year ago, prompting the restart of investments. SK Hynix will adopt a dual-track strategy: the Dalian factory will use Intel's mature floating gate architecture to produce NAND at the hundred-layer level, while domestic factories like Cheongju M17 will focus on advanced NAND production of over 300 layers, the latter having announced an investment of 19.1 trillion won. Industry insiders expect that the equipment configuration of Dalian Plant 2 will be similar to that of Plant 1, maintaining a monthly production capacity in the range of 40,000 to 60,000 wafers.

hot_img Intel's EMIB-T advanced packaging faces yield challenges, with the first mass production target in 2027 only at 50%

According to Taiwanese media reports, Intel's next-generation advanced packaging technology EMIB-T is facing severe yield challenges. ABF substrate supplier Unimicron stated during its earnings call on July 29 that the technology is "not yet mature," and the yield target for three EMIB-T substrate suppliers (Unimicron, Ibiden, and Shinkawa) during the initial mass production phase at the end of 2027 is only 50%.EMIB-T is Intel's advanced packaging solution that competes with TSMC's CoWoS-L, differing in that it embeds the silicon bridge directly into the substrate rather than using a silicon interposer, making substrate suppliers a key factor in determining success or failure. Google's ninth-generation TPU has confirmed that it will adopt EMIB-T packaging when it goes into mass production in 2028, with MediaTek responsible for chip co-design. If the chip successfully goes into mass production, its shipment volume is expected to challenge NVIDIA. Unimicron stated that the customer (Intel) has provided a profit guarantee mechanism, ensuring that suppliers can remain profitable even if yields are poor, and that profit margins will exceed the company's average level after achieving the 50% yield target. Currently, there is still considerable uncertainty about whether EMIB-T can achieve its mass production target by the end of 2027.
app_icon
ChainCatcher Building the Web3 world with innovations.