Micron Battles SK Hynix and Samsung for Engineers as HBM4 Production Surge Locks in $1,100 Analyst Target

Micron's talent war in Seoul highlights HBM demand surge. HBM4 mass production, full allocation through 2026, record revenue up 196%, and analyst targets up to $1,100 fuel 680% stock gain.

The memory chip boom has turned into a full-blown war for talent. Micron is placing job ads in Seoul explicitly targeting engineers at Samsung and SK Hynix, offering senior positions up to 300 million won — roughly $214,000 including bonuses. The company wants to build HBM expertise on the ground in South Korea without relocating foreign staff.

The hiring offensive underscores the ferocity of the current cycle. Micron’s fourth-generation high-bandwidth memory, HBM4, has entered mass production with a ramp that Executive Vice President Manish Bhatia says is running twice as fast as its predecessor HBM3E. Each module packs 36 GB of capacity and 2.8 TB/s of bandwidth, purpose-built for Nvidia’s forthcoming “Vera Rubin” architecture.

That speed is critical because every unit Micron can produce for the rest of the calendar year is already spoken for. The company confirmed that all HBM capacity through the end of 2026 is fully allocated to large customers. Broader shortages in DRAM, NAND, and HBM are expected to persist beyond 2026, with one analyst report pointing to tightness stretching into 2027.

Pricing power has followed. Contract prices for NAND and DRAM have surged 413% and 355% year-on-year respectively. For Micron, that translates into eye-popping financials. Revenue in the fiscal second quarter hit $23.86 billion, a 196% jump from $8.05 billion a year earlier. For the current quarter ending in May, management guided for about $33.5 billion in sales.

The cash flow picture is just as striking. On the JPMorgan technology conference stage, Micron signalled that the third fiscal quarter will deliver a record free cash flow — a milestone the company expects to hit even before the fiscal year wraps up. CEO Sanjay Mehrotra sees the broader AI infrastructure market swelling from $35 billion in 2025 to $100 billion by 2028, providing a multiyear tailwind for memory demand.

Analysts are racing to mark up their targets. Citigroup raised its price objective to $840. BofA Securities went further, calling for $950. Melius Research topped both with a $1,100 target, citing “unprecedented” demand from AI capital expenditure. Mizuho Securities, meanwhile, lifted its target to $800 with an “outperform” rating, pointing to the massive expansion of AI data centres.

The share price has followed the fundamentals. Micron stock trades around $656 after a 144% gain since the start of the year, hovering just below its all-time high. In pre-market trading, the shares added roughly 2% on the latest analyst upgrades.

New product launches are broadening the revenue base. The company is sampling low-power memory modules designed specifically for AI servers and expects to ship high-capacity data-centre SSDs in volume, with deliveries beginning in early May. As AI workloads shift toward inference applications, these products open up additional revenue streams beyond traditional DRAM components.

The stock’s 680% surge over the past twelve months reflects how far the narrative has moved. Yet the next catalyst is just weeks away. On June 24, Micron will report its fiscal third-quarter earnings, giving investors their first look at whether the pledged record free cash flow and $33.5 billion revenue target materialise.