NuAxi Stream - Real-Time Semiconductor Intelligence

Sun, May 3

22:53Impact score 92
Nvidia accelerates end-of-life for some Jetson AI processors due to memory shortages — RAMpocalypse sends older DDR4-based modules to the great scrapheap in the sky
Nvidia accelerated the end-of-life for older Jetson TX2 and Xavier modules to NCNR status due to LPDDR4 shortages. These legacy products (2017-2018) are being replaced by newer Orin/Thor systems using LPDDR5. Bullish for DDR4 pricing as supply tightens and legacy hardware becomes unavailable.
LPDDR4DDR4NvidiaTom's HardwareEOL/Discontinuation
20:51Impact score 15
Anthropic in early talks to buy DRAM-less AI inference chips from UK startup — Fractile's SRAM architecture reduces need for pricey memory during extreme pricing and shortage crunch
Anthropic is in early talks with Fractile for DRAM-less AI inference chips expected in 2027. Fractile claims 100x speedup over Nvidia GPUs to address Anthropic's high inference costs.
AI Inference ChipsSRAMDRAMAnthropic, Fractile and 5 totalTom's HardwarePartnership/M&A
20:08Impact score 85
Jensen says Nvidia now has 'zero percent' market share in China — says US export policy 'has already largely backfired'
Nvidia CEO Jensen Huang claims the company's AI accelerator market share in China has dropped to 0% due to US export restrictions. Bernstein previously estimated the share would fall from 66% to roughly 8%, driven by domestic vendors covering up to 80% of demand. This shift accelerates China's self-sufficiency drive, posing a long-term threat to US chip dominance in the region.
AI AcceleratorsNvidia, Huawei and 5 totalTom's HardwarePolicy/Regulation

Fri, May 1

21:29Impact score 15
La Luce Cristallina produces 200mm BaTiO₃ wafers at Austin fab
La Luce Cristallina has begun producing 200mm barium titanate (BaTiO₃) coated wafers at its new Austin fab, with an annual output of 200-250 wafers and plans for 300mm production. BaTiO₃ is positioned as a high-performance, foundry-compatible material for next-gen photonic ICs in telecom and AI data centers. Neutral: This is a niche R&D milestone with no immediate, quantifiable impact on mainstream component supply or pricing.
Photonics ICsOptical ModulatorsLa Luce CristallinaElectronics WeeklyProduct Launch
19:30Impact score 88
ASML's roadmap for chipmaking lithography tools examined — from DUV to Low-NA, High-NA, Hyper-NA, and beyond
ASML shipped 131 DUV tools in 2025, generating €32.7 billion in revenue. Chinese customers accounted for 70% of DUV sales in 2024 due to tightening Dutch export restrictions. Bullish for ASML and DUV equipment supply.
DUV lithography toolsEUV lithography systemsASML, TSMC and 6 totalTom's HardwareShortage Alert
15:01Impact score 88
Chip Industry Week in Review
2/3nm capacity crunch and memory shortage widening to 2027, alongside halted IC tool sales to China. Supply constraints from advanced nodes and export policies intensify. Bullish: Persistent supply-side pressure supports price uptrend.
2nm3nmMemoryTSMC, Samsung and 5 totalSemiconductor EngineeringShortage Alert

Thu, Apr 30

23:02Impact score 90
AWS says acute server memory shortage is driving customers to the cloud
AWS CEO Andy Jassy reports a server memory shortage is forcing enterprises to cloud. AI demand is diverting chip capacity to high-performance memory. This confirms supply constraints for DRAM/HBM.
DRAM HBMServer MemoryAWS, Google and 3 totalThe RegisterShortage Alert
22:46Impact score 95
Samsung and SK hynix warn AI-driven memory shortages could last until 2027 and beyond, as HBM demand explodes — customers already reserving supply years ahead, while the wider DRAM market begins to tighten
Samsung and SK hynix warn AI-driven memory shortages will persist until at least 2027, with fulfillment rates hitting record lows due to surging HBM demand. Manufacturers are reallocating capacity to high-margin AI memory, tightening supply for conventional DRAM and SSDs.
HBM DRAM SSDSamsung Electronics, SK hynixTom's HardwareShortage Alert
17:00Impact score 45
RT NEWS: 🚀 #MediaTek is shifting from a smartphone chipmaker to a AI #ASIC player, reportedly securing Google TPU v8 orders. It has doubled its 202...
MediaTek is reportedly shifting focus to AI ASICs and has secured orders for Google's TPU v8, doubling its 2026 revenue target for the segment to $2B. This strategic pivot leverages its foundry partnerships to capture growing AI accelerator demand.
利多
Potential long-term demand shift for MediaTek's advanced packaging and foundry capacity, but no immediate supply or price impact on components.
AI ASICMediaTekTrendForce Official (Market Data)Partnership/M&A

Wed, Apr 29

03:30Impact score 85
[News] U.S. Reportedly Moves to Block Chip Tool Shipments to China No.2 Foundry Hua Hong as It Seeks 7nm After SMIC
US Commerce Dept sent letters to Lam Research, AMAT, KLA blocking shipments to Hua Hong's 7nm fabs. Hua Hong targets 2026 output of several thousand 7nm wafers/month for Huawei AI chips. Bullish: Supply constraints tighten, increasing risk of shortage for AI chips and advanced equipment.
Hua Hong, Lam Research and 5 totalTrendForcePolicy/Regulation
03:17Impact score 25
Nvidia quietly launches 12GB RTX 5070 laptop GPU — midrange mobile gaming gets more VRAM amid the RAMpocalypse
Nvidia launched a 12GB RTX 5070 mobile GPU using 24Gb GDDR7 modules on a 128-bit bus, maintaining bandwidth parity with the 8GB version. This move addresses the memory shortage (RAMpocalypse) by offering higher capacity without bandwidth improvements. Neutral impact on spot prices as it targets a different capacity segment.
GDDR7 MemoryLaptop GPUNvidiaTom's HardwareProduct Launch

Tue, Apr 28

06:41Impact score 15
[News] CXL May Emerge as Post-HBM Battleground; Samsung Reportedly Presents System with 10× Performance Gain
Samsung presented a CXL 2.0-based memory system 'Pangea v2' at IEEE, claiming up to 10.2x performance gain over RDMA. SK hynix and Micron are also advancing CXL development, with Samsung planning a CXL 3.2-based 'Pangea v3' in 2026. Neutral: This is a long-term R&D roadmap with no immediate impact on component supply, pricing, or trading decisions.
CXL DRAMSamsung Electronics, SK hynix and 3 totalTrendForceTechnical

Mon, Apr 27

02:41Impact score 95
[News] Samsung’s May Strike Seen Disrupting Up to 4% of DRAM Output, With Weeks-Long Recovery Risk
Samsung Electronics' 18-day strike from May 21 to June 7 is projected to disrupt 3-4% of DRAM and 2-3% of NAND output, potentially causing 20-30 trillion won in losses and requiring weeks for recovery. Amidst tight global memory supply, this event involves 30,000-40,000 workers and threatens data center expansion plans for major cloud providers. Bullish impact: The supply shock exacerbates shortages, benefiting competitors like SK hynix and Micron.
DRAMNAND FlashSamsung ElectronicsTrendForceShortage Alert

Sun, Apr 26

20:00Impact score 60
Taiwan's stock market surpasses the UK's despite having less than a quarter of the UK's economy — AI boom propels Taiwan forward, TSMC alone accounts for more than 40% of Taiwan's total market value
TSMC reported record Q1 earnings with net income up 40.6% YoY to ~$18B and raised its full-year revenue growth guidance to >30%. The AI chip demand surge is driving market concentration, with TSMC now accounting for over 40% of Taiwan's total market value. Bullish: Strong earnings and raised guidance signal sustained high demand and pricing power for leading-edge foundry and memory suppliers.
AI ChipsHBMTSMC, Samsung Electronics and 3 totalTom's HardwareFinancial/Earnings

Sat, Apr 25

16:25Impact score 85
Intel Reports $13.6 Billion Q1 Revenue, Extends Earnings Beat Streak to Six Quarters
Intel reported Q1 revenue of $13.6 billion, beating guidance and extending its earnings beat streak to six quarters. Data Center and AI revenue grew 22% year-on-year, while foundry revenue increased 16%. Bullish: Strong AI demand and operational discipline drove the beat.
CPUXeon 6Core UltraIntelThe ElecFinancial/Earnings
05:25Impact score 10
DeepSeek's new models are so efficient they'll run on a toaster ... by which we mean Huawei's NPUs
DeepSeek released its V4 and V4-Pro open-weight LLMs, claiming benchmark performance rivaling top proprietary models and introducing architectural changes for improved inference efficiency. The models extend support for Huawei's Ascend AI accelerators. Neutral: This is a software/algorithm development with no direct, immediate impact on semiconductor component supply, demand, or pricing.
AI AcceleratorHuaweiThe RegisterTechnical

Fri, Apr 24

20:50Impact score 55
Intel bets the farm on AI inference to drag CPU back to the top table
Intel reported Q1 2026 revenue of $13.6B, beating expectations, with AI-driven business accounting for 60% of revenue, up 40% YoY. The company is betting on AI inference and agentic workloads to drive CPU demand, citing a favorable CPU-to-GPU ratio shift and recent wins like Xeon 6 in Nvidia's DGX systems.
利多
Intel's strategic pivot to AI inference and foundry progress (Intel 14A) could reshape mid-term demand for its CPUs and IPUs, but execution risk remains high.
CPUIPUIntel CorporationThe RegisterFinancial/Earnings
20:00Impact score 15
Meta Arms itself to the teeth by signing for 'tens of millions' of AWS Graviton cores
Meta has signed a multi-year deal to deploy tens of millions of AWS Graviton 5 CPU cores, continuing its diversification into Arm-based datacenter silicon which also includes Nvidia Grace/Vera CPUs and a custom Arm AGI CPU. This reflects a broader industry trend where proprietary Arm designs are gaining share in AI server infrastructure, potentially at the expense of x86 incumbents. Neutral: This is a long-term market share trend for Arm ecosystem vendors, but lacks immediate, quantifiable supply/demand impact on component pricing.
Arm-based Server CPUAmazon Web Services, Arm Ltd. and 5 totalThe RegisterMarket Analysis
19:09Impact score 95
Most Read – Optical transceivers, Price hikes, OCX scrapped
UMC has notified customers of proposed price increases in H2, following TSMC's announcement of multi-year hikes. TSMC expects 3-10% price increases for its advanced nodes (3nm, 4nm, 5nm). Bullish: Direct price pressure on foundry services, likely to cascade to downstream chip costs.
Foundry ServicesUnited Microelectronics Corporation, TSMCElectronics WeeklyPrice Movement
06:15Impact score 40
Intel CEO says he can 'think of no better partner than Elon Musk' to explore 'unconventional' ways to improve chip manufacturing — TeraFab partnership aims to rethink how chips are made to reduce costs
Intel CEO Lip-Bu Tan partners with Elon Musk's companies to license Intel's 14A node for TeraFab, aiming to reduce manufacturing costs through unconventional methods.
Intel, SpaceX and 3 totalTom's HardwarePartnership/M&A

Analog	$0.68	−5.9%
Connector	$0.75	+0.2%
CPU	$386.74	−21.3%
DDR4	$88.40	−0.2%
DDR52026-W13	$14.89	−11.2%
GPU	$478.20	+8.2%
MCU	$0.68	−61.7%Unusual
MLCC	$0.01	+54.0%Unusual
MOSFET	$0.12	−82.8%Unusual
PMIC2026-W13	$0.44	−12.5%
SSD/HDD	$288.43	n/a

NuAxi Stream - Real-Time Semiconductor Intelligence

Sun, May 3

Nvidia accelerates end-of-life for some Jetson AI processors due to memory shortages — RAMpocalypse sends older DDR4-based modules to the great scrapheap in the sky

Anthropic in early talks to buy DRAM-less AI inference chips from UK startup — Fractile's SRAM architecture reduces need for pricey memory during extreme pricing and shortage crunch

Jensen says Nvidia now has 'zero percent' market share in China — says US export policy 'has already largely backfired'

Fri, May 1

La Luce Cristallina produces 200mm BaTiO₃ wafers at Austin fab

ASML's roadmap for chipmaking lithography tools examined — from DUV to Low-NA, High-NA, Hyper-NA, and beyond

Chip Industry Week in Review

Thu, Apr 30

AWS says acute server memory shortage is driving customers to the cloud

Samsung and SK hynix warn AI-driven memory shortages could last until 2027 and beyond, as HBM demand explodes — customers already reserving supply years ahead, while the wider DRAM market begins to tighten

RT NEWS: 🚀 #MediaTek is shifting from a smartphone chipmaker to a AI #ASIC player, reportedly securing Google TPU v8 orders. It has doubled its 202...

Wed, Apr 29

[News] U.S. Reportedly Moves to Block Chip Tool Shipments to China No.2 Foundry Hua Hong as It Seeks 7nm After SMIC

Nvidia quietly launches 12GB RTX 5070 laptop GPU — midrange mobile gaming gets more VRAM amid the RAMpocalypse

Tue, Apr 28

[News] CXL May Emerge as Post-HBM Battleground; Samsung Reportedly Presents System with 10× Performance Gain

Mon, Apr 27

[News] Samsung’s May Strike Seen Disrupting Up to 4% of DRAM Output, With Weeks-Long Recovery Risk

Sun, Apr 26

Taiwan's stock market surpasses the UK's despite having less than a quarter of the UK's economy — AI boom propels Taiwan forward, TSMC alone accounts for more than 40% of Taiwan's total market value

Sat, Apr 25

Intel Reports $13.6 Billion Q1 Revenue, Extends Earnings Beat Streak to Six Quarters

DeepSeek's new models are so efficient they'll run on a toaster ... by which we mean Huawei's NPUs

Fri, Apr 24

Intel bets the farm on AI inference to drag CPU back to the top table

Meta Arms itself to the teeth by signing for 'tens of millions' of AWS Graviton cores

Most Read – Optical transceivers, Price hikes, OCX scrapped

Intel CEO says he can 'think of no better partner than Elon Musk' to explore 'unconventional' ways to improve chip manufacturing — TeraFab partnership aims to rethink how chips are made to reduce costs

Chip Teardown Lab2 opened

Component price indexNPI, vs. prior period

Trending components