Citi Global TMT Conf. 2025: Seagate CFO on Cloud Storage Demand, HAMR Ramp, and AI-Driven Growth

Key Takeaways

TL;DR: STX highlighted robust nearline HDD demand (+25% CAGR exabyte growth), fueled by GenAI/cloud, w/ exabyte vol. highly supply-constrained and scaling via HAMR. Margin/FCF tailwinds expected from cost structure and cap. discipline. Cyclical swings are dampened by order visibility and supply discipline. Near-term demand is anchored by multi-qtr build-to-orders and LTAs from top cloud customers; margin expansion as HAMR mix rises. No major cap. expansion planned; FCF prioritized for div. and buyback. Tariff risk monitored, but no near-term impact.

1. Demand Env. & Visibility

• "We see demand growth to be around mid-20s for nearline."
  • Nearline HDD demand: ~25% CAGR for exabytes.
  • Last 4Q: Nearline exabyte shipments +90% YoY, showing sharp recovery.
• Customer ordering
  • Orders structured for "3–4Q of visibility," direct build-to-order—reducing overstock/phantom inv.
  • "We need to have an order before we start the product": Demand volatility exposure reduced by real orders and DC refresh planning.
  • Multi-qtr LTAs, but mgmt. avoids extending visibility further due to customer forecast limits.
• AI & cloud as secular drivers
  • AI impact:
    1. Data Retention: Longer data retention for AI utility.
    2. Data Generation: AI creates new persistent data sets (esp. video/synthetic data like AV training).
    • "AI is extremely important for storage... now AI is generating data that is increasing storage. And I think this second part is probably the most important..."
  • Public cloud = ~75% of nearline vol., main growth engine.

2. HAMR Tech Ramp & Margin Leverage

• HAMR transition
  • "New qualification for our HAMR product. Now have four major cloud customers qualified."
  • 32TB HAMR ramp ongoing; 40TB HAMR qual. started July 2025 w/ major cloud customer.
  • Roadmap: Up to 100TB possible in current form-factor (10 disks, 20 heads).
  • Focus: "exabyte vol." (not units); exabyte share = HAMR mix/output, not chasing unit share.
• Tech impact on cost
  • "30TB HAMR drive already presents a slightly lower cost per terabyte than our last EAMR drive... 40TB is a much bigger decrease in cost per terabyte."
  • HAMR enables exabyte vol. growth w/o linear BOM increase: capacity ramps via higher areal density, not more disks/heads.
  • Material margin expansion: Mix shift to higher-cap drives structurally lowers $/TB, supporting higher margins.
• TCO
  • HDDs remain much cheaper per TB vs NAND; HAMR widens gap.
  • "TCO is still very much more attractive using HDDs... HAMR is a beginning, the cost reduction from one product to the next is much bigger..."

3. Competition, Tech Mix, & Regulatory/Tariff

• SSDs vs HDDs
  • SSDs/NAND = complementary (caching/compute); HDDs default for cold/persistent storage (“vast majority” of AI/back-end cloud on HDD).
• Tariff/Geo Risks
  • Two current HDD exemptions: Section 232 + USMCA; no short-term issues.
  • If lost, Thailand (main mfg.) faces 19% US tariff; main comps. similarly exposed.
  • US mfg. focused on heads (esp. HAMR) and R&D; could adjust footprint if needed, but no incremental CapEx expected absent regulation.

4. CapEx, FCF, & Capital Returns