How Apple Could Diversify Its Chip Supply Chain: Evaluating Samsung and Intel as Alternatives to TSMC
Introduction
Apple's reliance on TSMC for manufacturing its Apple Silicon chips has been a cornerstone of its hardware success, but the tech giant is exploring alternatives to reduce risk. Recent reports indicate that Apple has considered both Samsung and Intel as potential foundry partners, possibly as early as 2027 or 2028. While neither is as strong as TSMC in terms of process technology and capacity, the chip shortage and geopolitical pressures make diversification a strategic priority. This guide walks through the steps Apple (or any company) would take to evaluate and potentially adopt alternative chip manufacturers.
What You Need
- Deep understanding of current chip manufacturing ecosystem (TSMC, Samsung, Intel)
- Access to market research on foundry capabilities, yields, and timelines
- Data on each foundry's process nodes (e.g., 3nm, 2nm) and volume capacity
- Risk assessment frameworks for supply chain disruptions
- Internal team of semiconductor experts and supply chain analysts
- Long-term strategic planning horizon (5+ years)
Step-by-Step Guide
Step 1: Assess Current Dependence on TSMC
Apple’s A-series and M-series chips are currently exclusively made by TSMC. Start by quantifying the volume of wafers TSMC produces for Apple, the process nodes used (e.g., N3, N4), and the lead times. Recognize that TSMC offers the most advanced nodes with highest yields. This step highlights the risk: any disruption at TSMC could halt iPhone, iPad, and Mac production. For example, the 2021 chip shortage exposed vulnerabilities even for TSMC, although they managed better than others.
Step 2: Identify the Need for Diversification
Diversification isn’t just about having a backup; it’s about bargaining power and long-term security. Consider factors like:
- Geopolitical risks (Taiwan’s stability)
- Natural disaster risks (earthquakes in Taiwan)
- TSMC capacity constraints for future nodes
- Competitive pricing pressure
Apple’s own history shows that relying on a single supplier (e.g., Samsung for displays) caused issues. The company likely wants to avoid repeating that with chips.
Step 3: Evaluate Potential Alternative Foundries
Two main candidates: Samsung Foundry and Intel Foundry Services. Compare them based on:
- Process technology: Samsung has 3nm GAA (Gate-All-Around) but lower yields than TSMC. Intel is behind but aims for 2nm by 2025.
- Capacity: Samsung’s capacity is smaller and shared with its own products. Intel’s foundry is new and scaling up slowly.
- Track record: Samsung has produced chips for Qualcomm and others but had yield issues. Intel has never manufactured chips for external mobile/consumer SoCs at scale.
- Cost and business terms: Samsung and Intel may offer discounts to attract Apple, but TSMC’s reliability may outweigh lower prices.
Step 4: Consider the Timeline and Feasibility
According to Bloomberg, Apple has been considering these alternatives for some time, and a shift might happen by 2027 or 2028. That timeline aligns with when Intel expects its 18A (1.8nm) node to be mature, and Samsung to have 2nm ready. For each candidate, map out:
- When they can deliver the required node (Apple likely wants the latest)
- Volume ramp-up capability (can they supply tens of millions of chips?)
- Quality and yield consistency
Also factor in the chip shortage environment which adds urgency but also makes switching riskier because spare capacity is scarce.
Step 5: Evaluate the Weaknesses of Each Alternative
Both Samsung and Intel are “weak alternatives” compared to TSMC. Key weaknesses include:
- Samsung: Internal competition (Samsung LSI designs its own Exynos chips), lower yields, and less advanced EUV lithography maturity.
- Intel: Inexperience with large-scale external foundry customers, culture of designing its own chips, and delays in process node development (e.g., 7nm delays).
- Common issues: Both have smaller ecosystem support (EDA tools, IP libraries) compared to TSMC.
If Apple were to move a portion of production to them, it would likely be for older or less critical chips first (e.g., modems, IoT).
Step 6: Model the Risks and Rewards of Switching
Create scenarios with costs, timelines, and performance. For example, if Apple moves 20% of chip volume to Samsung in 2028, what is the impact on iPhone performance? Could it cause supply disruptions? Also consider the potential upside: leverage over TSMC in pricing and priority. The recent history of Apple considering buying Intel’s foundries shows that full acquisition was also on the table, but that idea didn’t materialize. Instead, Intel remains a foundry partner candidate.
Step 7: Make a Strategic Decision
Based on the evaluation, Apple will likely decide to:
- Maintain TSMC as primary supplier for leading-edge chips.
- Use Samsung for certain older or lower-volume products (e.g., Apple Watch, AirPods).
- Test Intel’s foundry for a specific chip (e.g., custom memory controller) as a pilot.
- Continue internal research into in-house manufacturing (unlikely).
The bottom line: diversification is necessary, but the alternatives are weak, so Apple will proceed cautiously, perhaps with a slow phase-in starting around 2027.
Tips for Success
- Don’t rush: Relying on subpar foundries can damage product reputation. Wait until yields match TSMC levels.
- Start small: Begin with non-core chips (e.g., Bluetooth, power management) before moving to main CPUs.
- Secure multiple sources: Having two backup foundries (e.g., Samsung and Intel) reduces risk further.
- Negotiate hard: Use the threat of switching to get better terms from TSMC, but don’t burn bridges.
- Monitor geopolitical developments: Taiwan relations will heavily influence the long-term viability of TSMC.
- Invest in supply chain redundancy: Just as Apple did with memory and displays, build relationships early.
In summary, Apple’s path to chip supply diversification is a careful balancing act between maintaining performance leadership and reducing dependence. While Samsung and Intel are considered, neither is a perfect replacement for TSMC – but as the chip industry evolves, they could become more viable by 2028.