Samsung delivered an all-time-record Q1 2026 (revenue KRW 133.9T +69% YoY; operating profit KRW 57.2T +756% YoY — single-quarter OP exceeded FY2025 full-year OP of KRW 43.6T), with HBM4/SOCAMM2 mass production for NVIDIA Vera Rubin commencing at $500-560 per stack and >80% gross margins. The stock rallied ~51% YTD-March-to-now (KRW 193,499 on 3/17 to KRW 292,500 5/22 close; intraday all-time high KRW 300,500 on 5/21) and ~10% in May alone (KRW 266,000 on 5/6 close), driven by three concurrent catalysts: (1) the May 20 11th-hour strike resolution averting the May 21-June 7 walkout, (2) Daishin/Korean-media reporting on an AMD 2nm Samsung-foundry engagement (note: contradicted for AMD's flagship EPYC line by AMD's own May 21 announcement that Venice ramps on TSMC 2nm and Verano will extend TSMC 2nm — any Samsung-AMD relationship is now restricted to lower-tier/client product at best, NOT EPYC/Ryzen flagships), and (3) Q1 results validating the AI memory super-cycle that TrendForce now projects to peak in 2027. However at KRW 286,000 (7/2 close; re-anchored 2026-07-03 from the 358,500 6/25 close after the July 1-2 memory-complex selloff round-tripped the entire June rally, -20.2%, back below the 292,500 5/22 scenario anchor): forward P/E ~14x on consensus 2026E EPS ~20,000 KRW is back in the middle of Samsung's 10-year range (6-20x) after touching the top at 358,500, and the SK Hynix HBM4 yield lead (1c DRAM ~80% vs Samsung sub-60% targeting 80% in H2) still caps Samsung at ~30% of NVIDIA Rubin allocation. The variant view: this is no longer a recovery-vs-cycle-peak debate (consensus has converged on multi-year super-cycle through 2027) — the new debate is execution. If Samsung closes the 1c yield gap by Q3 2026 and converts Qualcomm 2nm negotiations into a confirmed second flagship win (with the AMD-flagship narrative now invalidated, Qualcomm is the ONLY remaining 2nm flagship gating event), the framework's bull case is KRW 380K — now ~+33% above the live price after the pullback, and the stock trades ~16% BELOW the average analyst target ~KRW 339K (the June overshoot above consensus has fully unwound). The July pullback re-balances the risk-reward the 6/26 re-anchor had flagged as sharply downside-skewed — PW return recovers from -18.6% (at 358,500) to ~+2.0% at 286,000 — but does not yet reach the KRW 230-250K accumulate zone; if HBM4 yield disappoints or memory pricing rolls over in H2 2027, the asymmetric risk still skews bearish.