Pricing

A small die can cost more than a large one

The most common question is "what does this part cost", and the most common assumption is that price follows die size. In reality area feeds into only one of the three cost components, and even there it is just one of the multipliers.

Three things drive the cost

Each has its own driver. Total cost is the sum of all three.

  • 10,000× 5,000× Shots Scope area
    Area × magnification × layers

    Die imaging

    Driven by Process node, scope area and layer count

    Billed per shot

    Area does drive the shot count, but it is only one of the multipliers — and it is the area you outline, not the size of the die. A finer process needs higher magnification, and the field of view shrinks with the square of it: the same die shot at 10,000× takes four times as many frames as at 5,000×. Multiply by the number of layers to image, and that is the count.

  • 4 8 12 Al Cu Hours Metal layers
    Metal × layer count

    Delayering

    Driven by Metal composition (aluminium or copper) and layer count

    Billed per hour

    Independent of area — and of process node. What sets the effort is whether the metal stack is aluminium or copper, and how many layers there are. Copper is harder to remove than aluminium, and every additional layer compounds the alignment error, with chemistry and parameters re-tuned layer by layer.

  • Dense nets Simple nets Hours Device count
    Devices × connectivity

    Circuit analysis

    Driven by Circuit complexity — device count and connectivity

    Billed per hour

    What sets the effort is device count together with connectivity, and the two have to be read together. At the same device count, a densely laid out analog IC has far more nets to recover than a structurally simple part, and the hours can differ by an order of magnitude.

What sets the cost is not how big it is. It is how hard it is.

Which is why every project is evaluated first

We do not publish a price list. The variation in engineering effort between chips is too large for an unevaluated figure to serve either as your budget or as our schedule.

  • Two dies of identical size can differ several-fold in delayering time if one is six aluminium layers and the other twelve copper.
  • Two dies on the same process can differ by an order of magnitude in analysis effort if one is a power management IC and the other is dense analog.
  • Scope is a variable too. Analysing one critical block versus the full die does not scale proportionally.
  • Evaluating first is what makes the number meaningful — and what lets us commit to a real completion date.
A project schedule and a calculator used to cost each project individually
Every project is scoped and costed before a date is quoted — otherwise the number means nothing.

The evaluation report removes the uncertainty at no cost

Estimating cost is precisely what it exists for.

We need at least one sample. The evaluation report gives you die size, process node, metal and poly layer counts, and a top-layer photograph.

With the top-layer photograph in hand you can mark the region you want analysed, and we quote against exactly that. It is the least ambiguous way to agree scope.

The evaluation report is not engineering-verified and carries some margin of error, but it is sufficient to judge scale and feasibility.

The evaluation is free of charge. Getting that data means decapsulation, delayering and SEM time — we do the work and do not bill for it. You decide whether to proceed once you have the report and the price.

One analysis, two prices

One difference only: whether the report is ever supplied to anyone else. Who is reading it stays confidential either way.

Model Price Afterwards Best when
Non-Exclusive (default) Baseline (1×) CPS may list and sell the report to other customers you want the technology and a shorter development cycle, and do not mind the analysis being supplied to others later
Exclusive Never resold; held only by the commissioning party beyond the technology itself, you need certainty that competitors cannot obtain the same analysis
Rush

A rush order jumps the queue; it does not compress the work

If your schedule moves on you, a rush order at 3× the price skips the queue and enters the workflow at top priority. Engineering time still has a floor — delayering, imaging and schematic recovery each have to actually happen. A rush buys the front of the queue, not a shorter process, and we will not commit to a lead time that cannot be met; that kind of promise is no use to you.

About four in ten commissions choose Exclusive.

Where we draw the line

Whichever you choose, we never disclose who is reading a report — who commissioned the work, which chip they asked about, or when. This is not a paid upgrade; it is a line we have not crossed in eighteen years.

It does not have to be the whole die

Partial scopes are equally welcome, and cost follows scope.

Die imagery only

High-resolution layer-by-layer imagery with no schematic recovery. Cost here is driven mainly by the area you outline and the layer count.

Annotation only

Devices and connectivity annotated, without assembling hierarchical schematics. Suited to teams whose own engineers will interpret the result.

One critical block

Circle the region on the top-layer photograph and we analyse only that. Most first-time customers start here.

Common questions

Can you give a rough price range up front?
We would rather not. Any number given before seeing the die is either too low or too high, and both cause problems later. The evaluation itself is free of charge — it is the fastest route to an accurate figure, and there is nothing to lose by asking for one.
What if I cannot tell which block I need analysed?
Image first, outline second. Die imaging and circuit analysis can be commissioned separately: have us image the region you are interested in, and the delivered imagery can be stepped through layer by layer in BrigenOne. At that resolution the layout features are usually recognisable — the repetitive structure of memory arrays and APR digital blocks, the regular arrangement of analog cells. Once you can see them you can outline them. You mark the region against the imagery, and we quote the circuit analysis against that region, so neither side has to guess.
How long does it take?
Lead time depends on the same factors: process difficulty, scope and circuit complexity. Once we have your purchase order or signed LOA, we confirm an exact completion date along with the deliverables list.
How many samples do you need?
At least one for evaluation. The quantity required for the analysis itself depends on process difficulty and scope, and is confirmed with the evaluation.
Does the price include simulation?
No. We deliver schematics compatible with Cadence Composer, which you can convert to a netlist and simulate yourself, but simulation is outside the scope of this service.

Send us the part number

Or ship us a sample. After evaluation you get an accurate cost and schedule instead of a guess.