BTC $62,685.49 -1.47%
ETH $1,870.18 -0.83%
BNB $607.08 -0.65%
XRP $1.00 -0.97%
SOL $75.39 -0.71%
TRX $0.3333 -0.50%
DOGE $0.0694 -1.41%
ADA $0.1816 -1.97%
BCH $204.79 -4.69%
LINK $8.77 +0.90%
HYPE $56.39 -1.99%
AAVE $86.76 -2.45%
SUI $0.6768 -1.65%
XLM $0.1583 -1.36%
ZEC $486.34 -1.84%
BTC $62,685.49 -1.47%
ETH $1,870.18 -0.83%
BNB $607.08 -0.65%
XRP $1.00 -0.97%
SOL $75.39 -0.71%
TRX $0.3333 -0.50%
DOGE $0.0694 -1.41%
ADA $0.1816 -1.97%
BCH $204.79 -4.69%
LINK $8.77 +0.90%
HYPE $56.39 -1.99%
AAVE $86.76 -2.45%
SUI $0.6768 -1.65%
XLM $0.1583 -1.36%
ZEC $486.34 -1.84%

2027

All
Article
Flash

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.

hot_img SemiAnalysis: SpaceX may complete a 10GW data center by 2027, with expected inference revenue reaching $300 billion

Research institution SemiAnalysis released an analysis stating that SpaceX is expected to build approximately 10GW of AI data center capacity by the end of 2027. If 50% of this is used for inference services, with annual revenue exceeding $10 billion per GW, the annualized revenue could reach $300 billion. SpaceX CEO Elon Musk stated in the first earnings report that a "conservative estimate" suggests an additional 6-8GW will be added in 2027, with the actual figure possibly exceeding 10GW.SemiAnalysis's inference simulator shows that when running on the GB300 cluster at current startup cloud prices (about $3/GPU hour), leading model companies like OpenAI and Anthropic could generate annual inference revenue exceeding $10 billion per GW, with annual costs around $12 billion. Microsoft, with full access to OpenAI models and without bearing training costs, can also capture revenue of the same scale. The analysis points out that Microsoft has signed contracts for 10GW of data centers (total value exceeding $300 billion) since 2026, with a 90-day cancellation clause, significantly reducing signing risks.Regarding SpaceX's construction progress, SemiAnalysis believes Musk will significantly shorten the construction cycle by using onsite gas power generation, bypassing large power transformers, parallel construction, and shortening the debugging process. The Southaven plant in Tennessee expanded from 27 turbines (approximately 495MW) in February to 69 turbines (1.7GW) in July, and the "MiniHard" project can be completed in about 5 months with 450-500MW. However, the 10GW target still faces multiple challenges such as land approvals, gas supply, and equipment delivery. This analysis is based on model simulations, and actual implementation still carries uncertainties.

hot_img TrendForce: DRAM supply tightness will continue until 2027, Nvidia assesses and lowers Rubin Ultra HBM configuration

According to the latest research by TrendForce, the tight supply of DRAM is expected to continue until 2027, and the uncertainty of the HBM4e verification schedule is prompting AI chip manufacturers to consider reducing the HBM configuration of their next-generation products. Starting from Q3 2026, NVIDIA has begun evaluating the HBM solution for Rubin Ultra, expanding from the original 12-Hi HBM4e to alternative options such as 8-Hi HBM4e, 12-Hi HBM4, and 8-Hi HBM4, with the final specifications yet to be determined. Previously, due to the expected supply bottleneck of LPDDR5X lasting until 2027, NVIDIA decided to halve the SOCAMM capacity of the next-generation Vera Rubin Superchip module.TrendForce pointed out that the main reasons for the downgrade of Rubin Ultra from 12-Hi HBM4e include: the overall DRAM shortage in 2027 will limit the allocation of wafer capacity to HBM, and there is still uncertainty regarding the mass production schedule and yield ramp-up of 12-Hi HBM4e. It is expected that the HBM bit shipment volume in 2027 will increase by approximately 50-60% year-on-year, which will still be insufficient to meet the growth in demand. HBM suppliers will maintain pricing power throughout 2027, and HBM prices are expected to rise significantly. The final configuration will depend on the wafer allocation decisions of memory suppliers.

hot_img The industry claims that the DRAM capacity for 2027 has been sold out in advance, with HBM accounting for nearly 70% of DRAM capacity

According to DIGITIMES, industry insiders revealed that the three major memory manufacturers have completed the capacity allocation negotiations for the entire year of 2027 in advance, with both DRAM and HBM capacities sold out. In terms of NAND Flash, although it is not as tight as DRAM, it is expected that the capacity will also be fully booked by the end of August 2026. ADATA Chairman Chen Libai confirmed that applications related to HBM and AI servers will account for nearly 70% of DRAM capacity, while the supply of DRAM for PCs and mobile phones will be relatively limited.Industry assessments indicate that Samsung, Micron, and SanDisk's NAND capacity for the entire year of 2027 has also been pre-sold, with Kioxia and SK Hynix expected to confirm by the end of August. SK Group Chairman Chey Tae-won previously stated that the demand for AI semiconductors in 2027 is expected to increase by 60-100% compared to 2026, with storage demand expected to grow by 50-60%, leading to a potential widening of the supply-demand gap and possibly facing "the most severe shortage in history." Industry insiders pointed out that the capacity allocation model is more orderly compared to 2026, but the amount that manufacturers can provide is usually only 60-70% of customer demand, leaving some small and medium buyers still facing the dilemma of having no products to purchase. The price increase of memory may slow down in 2027, but remaining high could become the norm.
app_icon
ChainCatcher Building the Web3 world with innovations.