At a glance
- Per-die optical test before packaging requires passive alignment, trench-capable optical probes, and automated wafer handling on production wafer sizes.
- Glide-SiPh PIC aligns its optical probes passively using calibrated geometry and wafer registration, taking seconds per device with no optical-feedback search.
- InZiv builds the prober, probe card and fiber array in-house, coupling light into edge-coupled photonic integrated circuits passively.
- Omni-SiPh LightView maps light propagation and optical loss inside a working circuit when a wafer-test fail needs root cause.
- Glide-SiPh PIC and Omni-SiPh LightView are available to order; both target silicon photonics foundries moving from NPI toward volume.
InZiv
Published:
Choosing a wafer-level optical test platform for per-die data at foundry volume comes down to device access, measurement requirements and how the optical probe finds the coupling interface, because optical alignment can take most of the test time and vary from die to die. Glide-SiPh PIC, InZiv's automated tester for photonic integrated circuits (PICs), which are chips containing interconnected optical components, checks optical and electrical function while dies are still on the wafer. It aligns its optical probes passively, using calibrated geometry and wafer registration rather than an optical-feedback search at each device, so alignment takes seconds per device. InZiv's intended coverage spans test insertion points 1, 2 and 3: wafer testing before bonding (insertion point 1), wafer testing after bonding (insertion point 2), and testing of individual dies after wafer dicing (insertion point 3). For the companion problem, when a die fails and port measurements cannot say why, Omni-SiPh LightView uses near-field optical probes, meaning optical measurement at very short distances that can access the field extending just beyond a waveguide, to map light propagation and optical loss inside a working circuit at wafer and die level. Both systems are available to order. The sections below set out what per-die test data requires, how the available approaches compare, and what role each InZiv system plays in a wafer test flow.
What does per-die optical test data at wafer level actually require?
Per-die optical test data at wafer level means an optical result recorded for every device on the wafer, produced before packaging. That requirement constrains what gets measured and what hardware sits in the test cell.
What can a per-device test cover?
- Coupling and path loss. Insertion loss, the reduction in optical power caused by introducing a component or optical path, usually expressed in dB, is the base quantity. Coupling loss, the optical power lost when transferring light between a probe and a device, matters too, because process and yield engineers need to tell whether variation comes from the device or from the optical coupling.
- Spectral behavior. A taped-out device, such as a directional coupler, can come back with a measured split ratio that is off and shifts with wavelength.
- Electrical parameters of active elements. Depending on the device, this can include photodetector responsivity (output current per unit of incident optical power, in A/W at a specified wavelength) and dark current (current through a photodetector without illumination under specified conditions).
Glide-SiPh PIC's stated measurement set is multichannel optical and electrical characterization, optical insertion loss per channel, full-wafer functional pass/fail testing, and wafer-level yield and uniformity mapping.
Which equipment elements produce the data?
- Prober. Positions and handles the wafer for testing.
- Electrical probe card. Provides electrical contact between the tester and the device under test (DUT).
- Optical probe or FAU. A fiber array unit (FAU) holds multiple optical fibers for coupling light into or out of a device. An edge coupler couples light through the edge of a chip or an exposed facet within a wafer trench, while a grating coupler couples light through the surface.
- Alignment method. Active alignment uses measured optical feedback to find or maintain coupling. Passive alignment positions the probe from calibrated geometry and registration, without an optical-feedback search at each device.
- The customer's existing ATE. Automated test equipment from the ATE supplier provides electrical stimulus and measurement within the test configuration.
InZiv builds the prober, the probe card and the fiber array in-house.
Which wafer-level optical test approaches can a foundry buy today?
A foundry comparing wafer-level optical test approaches can weigh four established options against InZiv's passive-alignment system. Fixing the evaluation criteria first keeps the comparison grounded in the device.
- Alignment method. Whether coupling is found by an optical-feedback search at each device or set by calibrated positioning. This matters most when alignment takes most of the test time.
- Coupling access. Whether the optical probe can reach an edge coupler inside a trench, or couples through the surface at a grating coupler. New designs can put edge couplers in trenches too narrow for the current fiber array.
- Automation scope. Wafer loading, probe engagement, stepping and probe-card replacement determine how far the setup scales with volume.
- Data produced. Port-level pass/fail, per-channel insertion loss with wafer mapping, or spatially resolved measurement inside the circuit.
| Approach | Strengths stated in InZiv's framing |
|---|---|
| Active-alignment wafer probers | Large installed base, long photonics assembly and alignment heritage, single-sided silicon photonics systems already partly deployed, double-sided systems available today, probe-card scale, global service, established ATE integrations |
| Polymer-lens fiber arrays | Scale, a global supply chain, qualification with test-system integrators |
| In-house probe station with a hexapod | Already paid for, flexible for one-off experiments, familiar to its engineers, fine at low volume |
| Pass/fail testing at the ports | Costs nothing, changes nothing in the flow, good enough while volumes and yields are acceptable |
| Passive alignment (Glide-SiPh PIC) | Alignment in seconds per device without an active fiber search on every device; patented deep-trench FAU with pitch down to 30 µm; edge-coupled and grating-coupled devices |
Per InZiv's framing, teams that stay with their probe stations are usually at R&D or pilot volume, where an active search per device is slow but tolerable. InZiv's suitability for a given line depends on device access, measurement requirements, demonstrated performance and the customer's qualification schedule.
What roles do Glide-SiPh PIC and Omni-SiPh LightView play in a wafer test flow?
Glide-SiPh PIC and Omni-SiPh LightView play different roles in a silicon photonics wafer test flow: Glide-SiPh PIC is built for production testing, and Omni-SiPh LightView is built for analysis. Both are available to order.
Which test stages run before packaging?
Test stages before packaging are commonly numbered, and the numbering differs between sources, so each is named here in words:
- Insertion point 1, wafer testing before bonding. The photonic wafer is probed optically and electrically while whole, with light coupled through edge couplers in trenches or through surface grating couplers.
- Insertion point 2, wafer testing after bonding. The photonic wafer is bonded to an electronic integrated circuit (EIC) wafer. Depending on where the optical and electrical interfaces sit, this stage can require double-sided access, meaning probing from opposite faces of the stack.
- Insertion point 3, testing of individual dies after wafer dicing. Individual dies, such as optical engines, are measured once singulated.
InZiv's intended coverage spans these three stages. Testing of the finished co-packaged module is outside InZiv's fit today.
What does each system measure?
| Attribute | Glide-SiPh PIC | Omni-SiPh LightView |
|---|---|---|
| Role | Production testing: high-throughput pass/fail sorting at volume | Analysis: imaging light inside the circuit and identifying the root cause of a failure |
| Where it is used | High-volume production test; first use with a customer may be during the NPI ramp | Process development and NPI, including design validation, debugging yield problems and offline failure analysis |
| Level | On the wafer; 200 mm and 300 mm wafers, FOUP compatible | Wafer and die level |
| Coupling | Edge-coupled and grating-coupled devices | Edge and grating coupling |
| Measurements | Multichannel optical and electrical characterization, optical insertion loss per channel, full-wafer functional pass/fail, wafer-level yield and uniformity mapping | Light propagation and optical loss inside the working circuit, plus electro-optical measurements such as IV, LIV and wavelength |
| Probes | Patented deep-trench FAU, pitch down to 30 µm | Near-field, deep-trench and angled probes, with custom probe designs for specific customer architectures |
Glide-SiPh PIC's automation includes wafer handling (EFEM), probe-card replacement and probe alignment. Packaged modules, co-packaged optics assemblies and process parameters such as film thickness or critical dimensions are outside its scope.
Is an in-house probe station still good enough at production volume?
Whether an in-house probe station is still good enough at production volume depends on the setup and the volume. One setup is a paid-for probe station with a hexapod, where an active search locates coupling on each device; per InZiv's framing, teams that stay with their probe stations are usually at R&D or pilot volume. Another is pass/fail testing at the ports, which is good enough while volumes and yields are acceptable. A team may keep either because it meets current measurement and volume requirements, has already passed qualification, or integrates well with its preferred ATE and software.
| Do this | Watch out for this |
|---|---|
| Keep the hexapod station for one-off experiments | Optical alignment can take most of the test time and vary from die to die |
| Continue pass/fail at the ports while volumes and yields are acceptable | A yield problem can appear that port data cannot explain |
| Reuse existing probes on a new edge-coupler design | New designs can put edge couplers in trenches too narrow for the current fiber array |
| Maintain a custom system internally | Upkeep consumes engineering resources |
How should a foundry plan its evaluation and next steps?
A foundry can plan its evaluation and next steps for a wafer-level optical test platform as an ordered sequence, from stated requirements to an agreed configuration. Per InZiv, a purchase typically runs as follows:
- Technical discussion. Review the device architecture and test requirements, including the test insertion point in words (wafer testing before bonding, after bonding, or testing of individual dies after dicing) and the coupling interface (edge or grating coupler, and trench width where edge couplers sit in trenches).
- Sample-based demonstration or feasibility work.
- Configuration and acceptance criteria. Agree what the system must measure and how acceptance is judged.
- Commercial proposal and order. Followed by installation and acceptance. A pilot or on-site evaluation may be included and is optional.
Each role brings its own question. Test and metrology engineers look for accurate, repeatable measurements at the required throughput; process and yield engineers ask whether device variation can be separated from test variation; PIC design engineers ask whether the design provides the optical and electrical access needed; engineering and manufacturing leadership ask what has been demonstrated and whether it will run reliably in production; and procurement weighs price, delivery, warranty, service and supplier continuity. According to InZiv, the time from a demonstration to the purchase order is anywhere from 2 to 6 months and often depends on the customer's own timelines; that period ends at the order and does not include delivery or installation.
Frequently Asked Questions
What does passive alignment mean in a wafer-level optical test flow?
Passive alignment is optical positioning without an optical-feedback search at each device. In Glide-SiPh PIC, the fiber array unit is calibrated on a jig and placed on the wafer using fiducials, with no hexapod, no iterative tuning and no optical recalibration between cycles, and alignment takes seconds per device.
Which InZiv system sorts dies and which one explains a failure?
Glide-SiPh PIC is built for production testing, performing high-throughput pass/fail sorting at volume. Omni-SiPh LightView is built for analysis: near-field optical probes map light as it travels through the working circuit without damaging it. InZiv states that Omni-SiPh LightView is designed to reduce iteration cycles from weeks to hours, so a fail at wafer test can be traced to the bend, splitter, coupler or waveguide section where light is actually lost.
When does an in-house probe station remain the right tool?
A probe station with a hexapod is already paid for, flexible for one-off experiments, familiar to the engineers who run it and fine at low volume, where an active fiber search per device is slow but tolerable. Teams reconsider when alignment consumes too much of the device-test cycle, the optical probe cannot reach the coupling interface, coupling variation makes device results hard to interpret, volumes outgrow the setup's automation or throughput, maintaining a custom system consumes too much engineering, or a yield problem appears that port data cannot explain.
Does InZiv replace our existing automated test equipment?
No. InZiv's integrated scope is the prober, the electrical probe card and the optical fiber array unit, all built in-house. The customer's existing ATE, which supplies electrical stimulus and measurement functions, comes from the ATE supplier and remains part of the test configuration.
What falls outside InZiv's fit today?
Testing of the finished co-packaged module sits outside InZiv's fit today, as do very low-volume labs that only need input/output power measurements. Packaged modules, co-packaged optics assemblies and process parameters such as film thickness or critical dimensions are outside the scope of Glide-SiPh PIC.
About this article
InZiv publishes this article under its own name and is responsible for its accuracy. Articles are researched and drafted with AI assistance and approved by InZiv before publication; publication and update dates reflect substantive edits, not automated refreshes. Last updated: 2026-11-01