ASUS has stopped producing the NVIDIA RTX 5070 Ti and 5060 Ti 16GB, saying they've reached 'end of life'

ASUS has placed the NVIDIA GeForce RTX 5070 Ti and 5060 Ti 16GB models into end-of-life status and ceased production amid a constrained memory supply driven by surging AI-related demand for RAM and HBM, with retailers reporting shortages at least through Q1. NVIDIA maintains it is still shipping all GeForce SKUs and will lean on lower-VRAM 8GB models and the 12GB 5070, while higher-margin 5080/5090 lines are less affected; the shift tightens consumer GPU availability, could sustain higher component and retail prices, and pressures AIB partner product mix and margins.

Analysis

Market structure: Memory suppliers (MU, SK Hynix, Samsung) are the clear near-term winners as constrained DRAM/NAND/HBM capacity shifts volume and pricing power to datacenter/AI customers; expect DRAM spot tightness to persist through at least Q1 and likely into H1 2026, pressuring AIB margins by several hundred basis points and forcing OEM SKU rationalization (move from 16GB→8GB). Mid-tier consumer GPU SKUs (16GB RTX 5070 Ti/5060 Ti) and their AIB producers (ASUS et al.) are direct losers—reduced SKUs will tighten gaming supply, lift ASPs for available higher-margin 5080/5090 SKUs, and concentrate Volumes on 8–12GB parts.

Risk assessment: Tail risks include an accelerated re‑allocation of memory capacity to AI (massive capex shift), export controls on memory/AI tech, or a faster-than-expected rebound in consumer memory supply that collapses spot DRAM prices (>15% QoQ). Immediate (days) risks = retail stockouts and volatile secondary pricing; short-term (weeks–months) = AIB earnings hits/Q1 guidance misses; long-term (quarters–years) = structural capex tilt towards HBM reshaping supplier mix. Hidden dependency: rumor that NVIDIA stops supplying memory to partners would force AIBs into volatile spot markets and spike working capital needs.

Trade implications: Direct plays favor long memory suppliers (MU) and selective long NVDA exposure via time‑limited call spreads to capture secular AI while limiting premium bleed from consumer uncertainty. Pair trades: long MU vs short/underweight AIBs or consumer GPU–exposed names; options: buy 6–12 month MU calls and sell short-dated calls on AIBs or buy NVDA call spreads. Rebalance toward datacenter hardware and away from consumer-facing AIB inventory ahead of Q1 results.

Contrarian angles: Consensus ignores that SKU rationalization (fewer 16GB cards) could boost retail GPU ASPs and margin mix for NVIDIA, offsetting volume loss—NVDA may benefit net even as AIBs suffer. Historical parallel: 2017 crypto-driven GPU shortage temporarily inflated ASPs then normalized; if memory MSRPs collapse later in 2026, MU downside exists. Unintended consequence: sustained high component costs may retard PC upgrade cycles, capping long-term gaming GPU TAM and enabling console/streaming share gains for rivals like AMD.

Key Decisions for Investors

• Establish a 3% portfolio long position in MU within 5 trading days, horizon 6–12 months; target +30% upside, trim to half at +15%; cut if Micron guidance cuts revenue or gross margin by >200bps vs consensus or DRAM spot prices fall >15% QoQ.
• Buy a 9–12 month NVDA call spread sized 1–2% of portfolio (buy ATM call, sell ~15–25% OTM to fund) to capture secular AI upside while limiting premium; exit on 30% gain or if NVDA data‑center revenue guidance misses by >5% sequentially.
• Initiate a 0.5–1% short or buy a 3–6 month put spread on consumer/AIB‑exposed equity (eg. ASUS/board‑partner exposure or small position in AMD GPU segment) to hedge consumer GPU weakness; cover if company reports gross margin improvement >150bps or confirms stabilized memory procurement.
• Reduce direct holdings in consumer AIB manufacturers by ~50% ahead of Q1 earnings (redeploy proceeds into MU and NVDA LEAPs); re-evaluate after next two quarterly memory ASP prints or if NVIDIA re-commits to supplying memory to partners within 60 days.