Who added the most market value in 2026? 📊 Top 10 (Dec 31, 2025 → Oct 23, 2026): Apple +$1.12T (+28%) Micron +$670B (+208%) TSMC +$607B (+39%) Nvidia +$601B (+13%) AMD +$494B (+141%)

Samsung & SK Hynix are targeting Q2 2027 for mass production and volume shipments of 8-layer HBM4, specifically for NVIDIA’s Rubin Ultra rack-scale AI systems. While 12-layer HBM4 remains the primary focus through 2026, the share of 8-layer is expected to rise significantly from mid-2027 - driven by yield and thermal constraints on taller stacks. NVIDIA is adjusting its strategy and evaluating a mix of configurations. #HBM #HBM4 #NVIDIA #Samsung #SKHynix #AI #Semiconductors

NVIDIA Vera Rubin NVL72 is here. 72 Rubin GPUs + 36 Vera CPUs 20.7 TB HBM4 | 1,400 TB/s bandwidth 3,600 PFLOPS NVFP4 inference 2,520 PFLOPS NVFP4 training 288 GB HBM4 per GPU (19.2 TB/s) 6th-gen NVLink The next-gen AI rack is already in full production. $NVDA $MU $SKHY #HBM4 #VeraRubin #AI

NVIDIA just locked in lower HBM4 bandwidth for Vera Rubin.Earlier targets → Official specs (Oct 2026): • Rubin GPU: 22 → 19.2 TB/s • Superchip: ~44 → 38.5 TB/s • NVL72 Rack: ~1,580 → 1,400 TB/sCapacity stays 288 GB/GPU. Pin speed cut (~11.7 → ~9.5 Gbps) to improve yields and multi-supplier volume.Still a big jump over Blackwell’s HBM3E.#NVIDIA #VeraRubin #HBM4 #AIHardware

Amkor’s CapEx stayed flat ~$0.75–0.91B from 2021–2025.2026 guide jumps to $2.5–3.0B.NVIDIA is prepaying $2.8B ($1.5B Arizona + $1.3B Korea) to lock capacity.

SK hynix set for another monster quarter: 3Q26 consensus at KRW 98.2T revenue (+302% YoY) and KRW 77T OP (+576% YoY) with 78.4% margin. HBM4 Vera Rubin sales kicking in, DRAM ASP +21-22% QoQ. Will HBM4 shipments beat expectations on Oct 28? #SKhynix #HBM #Memory

NVIDIA, AMD & Broadcom have locked in massive supply chain commitments. Here’s how those obligations stack up against annualized revenue, net income, and free cash flow.

Micron targets 2028 for Vertical HBM (vHBM) - memory stacked directly on the processor. • HBM4: 2.8 TB/s per stack • ~100 stacks ≈ world’s entire internet bandwidth (282 TB/s) • vEIM 1: 2 TB/s @ <1 pJ/bit Standard, Custom & Vertical HBM will coexist.

Boldest analyst calls this week 📊 $SPCX JPM $500 (+239%) $STX BofA $1,200 (+90%) $MU Needham $1,650 (+88%) $NVDA GS $300 (+42%) $TSLA Piper $475 (+27.6%) $CRWV HSBC $39 (−28%) #Stocks #WallStreetJournal

Huaqiangbei RTX 50-series channel quotes (Oct 23):5090 → 55-60k RMB (from 15-16k launch) 5080 → 16-17k 5060 → 3,600+Prices are jumping daily. Via 21st Century Business Herald. #NVIDIA #RTX50

AI Component Lead Times: Current vs Equilibrium Baseline HBM, High-performance ASICs, AI Accelerators, 800G Optical Modules, and HBM Interfaces are running at roughly 2× their normal (balanced) lead times due to AI demand. Current figures (blue) vs equilibrium baseline (green). Unit: weeks $HBM

Japan’s semiconductor equipment billings hit $12.0B in Q3 2026 - up 22.5% quarter-over-quarter and 48.2% year-over-year.AI-driven demand for advanced logic and HBM continues to drive strong capex across the supply chain.$8035.T $6857.T #Semiconductors #AI #TokyoElectron #Advantest

Server CPU market is shifting. TrendForce estimates show Intel’s share falling from ~50% in 2026 to 42% in 2027, while Arm-based CPUs (NVIDIA, AWS, Google and others) continue to gain ground. AMD holds steady near 28%. #ServerCPU #AI #Intel #AMD #NVIDIA #Arm #DataCenter

SK hynix at 2.8x forward P/E. Samsung at 3.5x. Micron at 5.4x. CXMT at 10.3x. The Korean memory majors are cheaper than they’ve been in past downcycles. CXMT is the expensive one.

Estimated Server CPU Market Shares: 2026 vs 2027 (TrendForce) Intel’s lead shrinks while AMD holds steady and custom silicon (AWS, Google) gains ground. #ServerCPU #Intel #AMD #NVIDIA #TrendForce #Semiconductors

Global memory makers have had a massive run - despite recent pullbacks from peaks.SNDK +826% (1Yr), Kioxia +451%, Micron +338%. Even after the drawdowns, YTD returns remain extremely strong across the board.

Nomura estimates: Global data center power by applicationAI goes from 10GW (17%) in 2025 → 40GW (35%) by 2028Heavy agents become the biggest AI power user (15GW), ahead of training (10GW)

Memory stocks trading at just 2–3x earnings.The market is already pricing in a 2027 downcycle that still hasn’t shown up in the fundamentals.Samsung 2027: current 3.2x vs DCF-implied 9.1x.

Micron just laid out its full memory roadmap through 2030. • DDR6: >2× bandwidth vs DDR5 • LPDDR6: >35% better power efficiency (1δ node) • HBM5 (2028) → HBM5E (2029) → HBM6 (2030) • vHBM with hybrid bonding starting 2028 • G11 NAND: >400 layers, 2× density vs G9, ramp in 2H27 Vertical stacking is coming. The memory wall is real. #Micron #HBM #DDR6 #NAND #AI #Semiconductors $MU

Micron just laid out its AI memory roadmap through 2030. Token efficiency = Bandwidth + Energy Efficiency + Capacity • HBM (JEDEC + Custom) → HBM4 to HBM6 • vHBM (Vertical 3D HBM) starts 2028 • vEIM (Vertical Edge Inference Memory) targets 2TB/s & <1pJ/bit Vertical stacking is coming. Memory wall is real. #HBM #Micron #AI #Semiconductors #vHBM