Copyright © 2020-2021 Shenzhen CDTech Electronics LTD. All rights reserved. Site Map Powered by iwonder.cn
display / touch / bonding solutions
display / touch / bonding solutions
A purchase order for 5,000 modules and a launch plan reaching 40,000 a month by the third quarter are two different documents. The first is a commitment; the second is a forecast about a future nobody has built. Most ramp failures start when a planning team sends the second and assumes it secured the first.
Display capacity is physical: a line slot, a glass allocation, an IC position, test fixtures, a trained team. Reserving those costs a supplier money in declined work, so none reserves them against a spreadsheet alone.
This guide separates the reservation from the commitment, then shows the four numbers that turn one into the other.
A ramp plan reserves capacity; only a call-off buys units.
The plan obliges a supplier to hold line time, glass and material; the purchase commitment starts when a dated call-off enters the agreed firm window.
Confusion between the two is mostly vocabulary. A forecast states intent, a reservation agrees availability, a purchase order obliges payment, and each carries a different remedy. A planner who emails a twelve-month build plan has sent a demand signal; a supplier who answers that it can support the volume has said the volume is plausible, not that the line is booked.
A reservation sits between those documents: monthly capacity, a firm window, a cancellation ladder, a price basis, and usually a fee or minimum take. Without them, booked capacity is a conversation. Comparing what suppliers charge for that arrangement is covered in comparing display supplier quotes; this page starts where one supplier is chosen and the volume gets built.
A supplier cannot reserve capacity against an adjective. "Ramping fast" and "high volume" translate into nothing on a production floor. Four numbers do.
| Number | Unit | What it fixes | What fails without it |
|---|---|---|---|
| Peak month volume | units per month | Line, glass and IC allocation at the top | Capacity sized to launch, not peak |
| Ramp slope | months to peak | Tooling, fixture and second-shift timing | Capacity arrives after the demand |
| Variant split | peak share per part number | Changeover, fixtures, material positions | One variant eats the reservation |
| Firm window | weeks non-cancellable | Share of reservation that binds | Forecast read as an order |
The set matters more than the order. Peak volume without a slope says how high the ramp goes but not how fast. Variant split, the number most often missing, decides whether the line runs one product or six. The firm window turns the other three into something schedulable.
The forecast horizon must be longer than the longest item bought or built before the first unit ships. On a custom display that item is rarely the glass; it is usually tooling and qualification: a backlight mold, an FPC tooling set, a bonding fixture, a test jig and the sample rounds that prove the assembly.
Where qualification runs longer than the forecast window, the forecast cannot protect the ramp. A plan that rolls thirteen weeks while tooling and sample approval take two quarters has a horizon ending before the constraint begins, and the ramp then looks like a supplier failure when it was arithmetic.
Two horizons work better than one. Monthly buckets extend far enough to cover tooling, material and qualification, even at low confidence. Weekly buckets carry the binding dates. A rolling forecast for an LCD program is therefore a release schedule for the longest lead item with the demand curve attached, not a consumer demand plan.
The buyer's call-off binds; the forecast never does.
Inside the firm window a call-off is a purchase obligation; outside it, the forecast only lets the supplier buy long-lead material at an agreed risk.
The mechanics decide who pays when plans move, so they belong in writing. A schedule has three zones: a forecast zone from the end of the firm window to the horizon, advisory; a firm zone, binding in volume and date; and a ship zone, where only logistics can still change.
Long-lead material complicates the middle zone. Glass, driver ICs and custom backlights are often ordered before the firm window reaches them, so someone carries the exposure. Either the supplier buys against the forecast and prices the risk, or the buyer releases material authorization with a liability cap. The delivery point belongs in the same document: the Incoterms 2020 rules matter because a call-off naming a quantity and a date but no delivery term is half a commitment.
A late ramp moves volume; the reservation still stands.
The reservation holds, so the supplier keeps the capacity and the buyer absorbs the material already released against the original ship dates, including any minimum lot already in production.
Put the adjustment in the schedule agreement and most of the argument disappears. Apply it to the next call-off, spread over the remaining months:
Adjusted call-off (month n+1) = planned call-off + (cumulative deviation / recovery months)
Cumulative deviation is actual minus planned shipments, recovery months are those left in the window, and the result is capped at reserved monthly capacity. A late ramp gives a negative deviation, so the schedule pulls volume forward or carries material into later buckets. If the slip outlasts the window, renegotiate the window rather than accumulate a carry-forward no line can absorb.
An early ramp is harder: unreserved capacity rarely appears on request, so a pull-in can be confirmed only within the agreement's upside band and against material already on order. That band, easy to treat as boilerplate, is often the most valuable clause in the peak month.
Two numbers that look alike cause most of the friction in a first ramp. Minimum order quantity is commercial: the smallest order a supplier will accept at a given price. Minimum production lot is physical: the smallest run a line can start without wasting material.
On a display module the lot is shaped by the glass sheet, the minimum reel on the SMT line, the backlight batch and the test jig. A supplier can accept an order of 1,000 units while running a 3,000-unit lot, because the surplus is absorbed by the next call-off; it cannot run below the process minimum without charging for the waste. Each lot also carries the material declarations the destination expects, such as RoHS 2011/65/EU evidence.
The first two or three call-offs therefore cost the most per unit, and price steps at lot boundaries.
Allocation in a tight market follows commitment, not conversation. When glass or driver IC supply tightens, suppliers rank customers on firm orders inside the window, willingness to carry material liability, payment record and forecast accuracy. A planner with six months of clean schedules has more leverage than one who escalates loudly once.
The path should exist before it is needed: a planner-to-planner review with the schedule as evidence, then an account-level commercial review where quantities can be traded between programs, then an executive call, worth spending only when the first two produced a documented gap.
Bring the firm orders, an offer to carry long-lead material liability, and a flexibility statement. The last is often decisive: an approved alternate driver or backlight can free capacity that a redesign cannot reach in time. Regulated chains add a layer: quality and change-control records in the AIAG framework are part of what a supplier protects when it moves one customer's volume ahead of another's.
Most schedule disputes are data disputes wearing a commercial costume. Two teams can agree on the plan and still disagree about what it says, because their files define month, week, part number and quantity differently. One versioned format, published on a fixed day, removes that category of argument.
Reconciliation matters more than the format. The schedule should tie back to open purchase orders and to the last shipment that left, so cumulative deviation is visible rather than discovered at quarter end. Keeping the same numbers from first sample to end of life is easier where continuity is designed as a lifecycle problem, as the supply continuity guide sets out.
A capacity reservation buys the option to build, and an option has a price. The trap is paying for the whole ramp whether or not the market arrives. Structures that avoid it let commitment grow as demand firms.
A fee credited against future orders and released in tranches pays for held capacity while exposure stays tied to demand. A ratchet, where the committed percentage stands only if demand materializes, does the same in contract language.
Vet the terms before signing:
The questions bite hardest on custom tooling, where the supplier invests ahead of any order. A draft schedule and draft terms can be reviewed through the contact page, and EU ecodesign requirements for electronic displays belong in the same review, since a compliance gap can stop a committed build from shipping. The sequence leading there is set out in the custom sourcing guide.
Next step. If your team is preparing a ramp and needs a capacity number rather than a range, send the four inputs above and the target months. CDTech can reply with a capacity position, the lot constraints behind it, and a draft call-off structure to review together.
Is a forecast a commitment?
No. A forecast is an advisory statement of expected demand and carries no purchase obligation on either side. The commitment starts only when a call-off falls inside the firm window agreed in the schedule, and only for the quantities and dates that window covers. Suppliers do buy long-lead material against forecasts, but that exposure is either priced into the quote or covered by a separate material authorization.
How far ahead must capacity be booked?
Far enough to cover the longest item that has to be tooled, bought or qualified before the peak month. Work backwards from the peak shipment date through tooling, sample approval, material lead time and production; the earliest of those dates is when the reservation has to exist. If that date already sits inside a forecast you have not sent, the ramp is running on the supplier's goodwill rather than on a plan.
What happens if the ramp ships late?
The reservation usually survives the delay and the volume moves. Material released against the original dates has to be absorbed, either as a carry-forward into later buckets or as stock held for the program. If the slip runs beyond the reservation window, the window itself is renegotiated, because a carry-forward larger than the line can produce in the months remaining is not a plan. Agree the adjustment formula before the delay, not after it.
Does a call-off schedule affect unit price?
It can, through three channels. Small early call-offs below a lot multiple carry setup cost. Volume spread evenly across the year is priced better than a sharp peak, because the supplier does not have to hold idle capacity for the rest of it. And a long firm window lets the supplier buy material on better terms, sometimes shared back. The schedule is a price input, not only a logistics one.
By continuing to use the site you agree to our privacy policy Terms and Conditions.