Micron Technology and Sandisk enter a pivotal week with two events likely to set the tone for memory-chip and AI-infrastructure sentiment: Micron's Taoyuan labor talks on Tuesday 27 October and Sandisk's fiscal Q1 FY27 earnings on Thursday 29 October after the close.
Micron closed Friday at $877.55, up 2.46%, after Thursday's 6.3% drop, with after-hours quotes near $879.90. The Taoyuan union rejected the company's latest offer by nearly 70% of voting members, with 1,633 members voting against. The union demands 15% of global operating profit paid as quarterly bonuses and a one-time bonus of 83 months' salary; Micron's earlier offer was up to 68 months, plus a new quarterly bonus plan from fiscal Q1 2027. Eleven union officials held a 24-hour warning strike from 22 October, but no full walkout has been called. Talks were moved from Friday to Tuesday because of Typhoon Champi.
SK hynix's third-quarter report, due before 9 a.m. Seoul time on Wednesday 28 October, is the first hard DRAM number of the season. Its shares rose as much as 2.6% on Monday after SK hynix shelved the US IPO review for its NAND and SSD subsidiary Solidigm. The decision removed a capital-structure overhang, though South Korea's financial regulator had warned about tightened dual-listing rules. Kioxia, Sandisk's flash partner, gained about 1.2% in Tokyo, offering a closer read on NAND sentiment.
Sandisk closed Friday at $1,238.03, down 1.82%, about 47% below its 52-week high. The company guided to non-GAAP EPS of $44-$46 on revenue of $10.3-$10.8 billion, but consensus estimates sit at or above the top of guidance, with revenue near $10.64 billion. Options markets imply an 8%-14% move around the report. Analyst targets range from BNP Paribas Exane's $1,400 to Bernstein's $3,000.
Elon Musk added a new risk over the weekend, saying SpaceX closed $40 billion of debt for GPU purchases and that debt markets have "almost entirely shut off access to money" for AI-only data-center operators. The warning pressured chip sentiment in Seoul and highlights dependence of memory pricing power on AI data-center demand. For crypto markets, the read-through is indirect: the news may influence macro risk appetite and AI-infrastructure funding sentiment, but there is no direct token-level catalyst.
