FRACTAL INFINITYAnalysis Tools
Bottleneck & Supply Chain X-Ray
Find what actually prevents a product, project, or industry from scaling — then discover what becomes the next constraint when that bottleneck is removed.
Most industries do not stop growing because demand disappears. They hit physical limits somewhere inside the system: a component with too few qualified suppliers, a manufacturing process that cannot expand quickly enough, a specialized material, limited equipment throughput, difficult qualification, low yield, or infrastructure that takes years to build.
The X-Ray is designed to find those constraints.
Instead of looking only at the companies everyone knows, Fractal decomposes a system through the physical chain underneath it:
Product → Component → Process → Equipment → Material → Supplier → Capacity → Qualification
It then uses sourced evidence to examine where demand is colliding with real-world capacity, how difficult that constraint is to substitute, how important it is downstream, and what could relieve it.
The result is not simply a list of supply-chain risks. It is a map of what controls scale.
What can an X-Ray reveal?
An X-Ray can help answer questions such as:
- What breaks first?
- If production increases 2×, 5× or 10×, which dependency becomes unable to keep up?
- What's underneath the obvious bottleneck?
- If advanced packaging is constrained, is the true control point packaging floor space — or substrates, materials, bonding equipment, yield, test capacity or qualification?
- What happens if we fix it?
- Relieve the current constraint and Fractal recomputes the system to identify the next likely bottleneck.
- Can we route around it?
- Alternative suppliers and technologies may exist technically but still require redesign, qualification or years of capacity expansion. X-Ray distinguishes an apparent substitute from a practical one.
- Where could the market be looking in the wrong place?
- A relatively small material, equipment supplier or manufacturing process can control production worth orders of magnitude more downstream.
How the X-Ray works
-
Define the system
Choose what you want to understand and the scale being tested.
Example: AI accelerator production → 5× growth
Future investigations can apply the same instrument to datacenters, humanoid robots, nuclear infrastructure, aerospace systems and other physical industries.
-
Decompose the physical system
Fractal traces the dependencies underneath the finished product instead of stopping at the obvious suppliers.
-
Ground the model in evidence
Important assumptions are connected to observable evidence such as manufacturer disclosures, earnings calls, technical publications, capacity announcements and other attributable sources.
Fractal distinguishes between:
- Observed — supported by evidence
- Modeled — an explicit engineering judgment
- Calculated — output produced by the X-Ray
-
Identify the binding constraints
The system evaluates factors such as demand pressure, scarcity, supplier concentration, lead-time pressure, substitution difficulty and downstream importance.
Evidence confidence is shown separately from constraint severity. A severe bottleneck with weak evidence is not presented as equivalent to one supported by strong evidence.
-
Look underneath the bottleneck
Finding that “CoWoS is constrained” is useful. Finding what prevents additional CoWoS capacity from becoming available may be considerably more valuable.
X-Ray therefore looks for both:
- Observed Bottleneck — what the market currently cannot obtain fast enough.
- Underlying Control Constraint — what prevents industry from relieving that shortage.
-
Relieve it and continue
The investigation does not stop with today’s bottleneck. Fractal asks: If this constraint disappeared tomorrow, what would break next?
That produces a constraint cascade — a view of how bottlenecks may migrate as technology, manufacturing capacity and demand change.
Read the methodology
How does Fractal decide something is a bottleneck?
Read the X-Ray technical paper for the architecture, evidence standards, modeling assumptions, scoring methodology, constraint-relief logic, provenance rules and limitations behind the instrument.