【DRAM Spot Buying Stays Cautious; DDR4 2Gx8 Prices Decline】 In the DRAM spot market, procurement momentum has remained cautious and watchful this week, with price trends diverging across segments. Spot prices of DDR4 2Gx8 branded chips have corrected downward noticeably, yet no clear buying interest has emerged in the market. Price trends of DDR5 products have varied by specification. Prices of eTT chips have been relatively strong as transaction prices continue to climb. However, buyers’ willingness to chase higher prices has been dampened by sharp quote hikes and suppliers holding back inventory, limiting the pace of actual market follow-through. The average spot price of mainstream chips (i.e., DDR4 1Gx8 3200MT/s) has increased by 0.55% from US$45.79 last week (September 16) to US$46.04 this week (September 22). 📷 Related news: 🔗 📷 For more on #DRAM# Price: 🔗

#AMD# has joined the $1T club, and its #CPU# business could get another boost. As #Meta’s# Muse gains traction, personal AI agents may drive greater #CPU# demand, benefiting the chipmaker alongside #Intel# and Arm.💡More: 🔗

📺 Strong AI demand is squeezing TV panel material capacity, pushing panel costs up an estimated 4–7% QoQ in 4Q26 while demand declines. 💡More analysis from #TrendForce#: 🔗

U.S. CSPs are ramping up #eSSD# orders, with demand rising over the past two weeks. QLC products are seeing the strongest growth, with penetration continuing to increase.💡More analysis from #TrendForce#: 🔗

🦾 Humanoid Robot Hands: The Costliest Battle in Physical AI. Dexterous hands are the most expensive module in humanoid robots. Three technical routes are emerging: screw-linkage, tendon-biomimetic, and high-perception sensing. Tesla pursues in-house precision manufacturing, though mass-production readiness remains unproven. 1X targets homes with tactile tendon designs. Who wins the dexterous hand race 👉 #TrendForce# #SelectedTopics#

#Apple# is reportedly cutting jobs again, this time affecting Fitness+ audio roles. The move follows 200+ layoffs in August across its Vision Pro, Siri, and AI teams. 💡More: 🔗

China’s #MLCC# market is splitting: AI demand is keeping some high-end orders booked into 2Q27, while consumer spot prices are sliding sharply from earlier highs amid weak demand. 💡More: 🔗

#TSMC# is speeding up its #CoWoS# push with a new “Mini-Loop” at Kaohsiung’s Baipu Industrial Park, aiming to make equipment validation 25–50% more efficient before tools enter production. 💡More: 🔗

Chinese researchers reportedly developed a new #Quantum# Random Access Memory (QRAM) router that simplifies complex routing to cut circuit depth and errors. Tests showed up to 98% transmission efficiency.💡More: 🔗

Is the long-awaited #Optimus# Gen 3 finally nearing production? A #Tesla# team reportedly landed in Ningbo last week, kicking off fresh supplier audits in China involving longtime auto partners including Tuopu Group, Joyson Electronics and more.💡More: 🔗

MLCC industry leader signals capacity expansion! Executive reveals optimism about AI-driven demand, describing the current assessment as “more ambitious.” 📰 News Source:

As AI chip sizes grow and integrate more HBM, 2.5D packages are getting larger, straining TSMC’s CoWoS capacity. Could Intel’s lower-cost EMIB-T shift CSP orders? 💡More analysis from #TrendForce#: 🔗

⚡ EVs Are No Longer Just Green. They're Energy Security. The US-Iran conflict blockaded Hormuz and spiked global oil prices, exposing the fragility of fossil-fuel transport. NEVs have shifted from carbon-reduction tools to strategic energy-hedging assets. Reduced oil dependency now directly translates to national resilience against geopolitical shocks. Explore the new EV supply chain landscape 👉 #TrendForce# #SelectedTopics#

The Chinese Academy of Sciences reportedly developed GAA transistors using #DUV#, potentially paving a path toward 3nm-level performance without relying on EUV lithography.💡More: 🔗

China's #DRAM# leader #CXMT# may be eyeing a push into #NAND#, potentially taking on #Samsung#, #SKhynix# and #YMTC#. Talks with potential customers, including an AI startup, are reportedly underway.💡More: 🔗

#Apple# is pushing #iPhone# cooling further with the 18 Pro, featuring new A20 Pro packaging and a 3× larger vapor chamber for up to 40% higher sustained performance. 💡More: 🔗

A 56-week lead time sounds long. Long compared to what? For ABF substrates, a balanced market is closer to 12 weeks. Current lead times are running at 48–56 weeks — more than 4x normal levels. We track current vs. balanced lead times across six AI infrastructure components, every Monday. (1/2)