Most shops choose a granite cutting machine by looking at the biggest slab they might one day cut, then buying the smallest machine that can handle it. That reasoning produces two failures: a machine that cannot make the cuts the shop actually sells, and a machine that can, but that nobody on the floor knows how to program.
The decision is not “which machine”. It is “which cut list”. Write down the last fifty jobs you quoted — the slab sizes, the shapes, the edges, the cutouts — and the machine selection mostly makes itself. This guide is about how to do that, and how to verify what you are buying.
What a granite cutting machine has to do
Four jobs get bundled under “cutting”, and they have different mechanical requirements:

- Dimensional cuts. Straight cuts from a slab to a finished countertop or tile size. Needs accuracy and repeatability more than flexibility.
- Cutouts. Sink openings, hob openings, tap holes, socket cut-outs. Needs a tool that can enter the material away from the edge and travel in two dimensions inside the slab.
- Edge work. Mitres, bevels and profiles. Needs a head that can tilt, or a separate machine.
- Repetitive strip cutting. Many identical pieces from many slabs. Needs throughput and low setup time per piece, not shape capability.
A machine built for one of these is efficient at it and compromised at the others. Most shops that struggle with cutting have bought one machine and asked it to do all four.
The five machine families
| Family | What it is good at | What it cannot do | Buy it if |
|---|---|---|---|
| 3-axis CNC bridge saw | Straight dimensional cuts, step cuts, basic drilling with a cutter installed | Mitres without re-fixturing; shaped cutouts | Your work is dimensional countertops and tiles |
| 4+1 axis bridge saw | Straight cuts plus mitres and bevels from one setup, because the head rotates | Complex internal shapes | You sell mitred edges and want them off the same machine |
| 5+2 axis bridge saw | Straight, mitred, shaped and internal cutouts; arcs and profiles by interpolation | High-volume strip cutting | You cut sinks, curves and mitres on varied slabs |
| Multi-blade strip machine | Many identical strips at once | Anything that is not a straight strip | You produce skirting, borders, mosaic strips in volume |
| Waterjet / sawjet | Any shape in any hardness, tight nesting, no heat | Speed on plain straight cuts; running cost per metre | Your value is in shape complexity, not in throughput |
The axis numbers are not marketing decoration. They describe how many directions the cutting head can be positioned and orientated, and each one removes a manual operation or a second machine from the process.
HOYUN’s own range covers the first four positions — the 3-axis CNC bridge saw, the 4+1 axis machine, the 5+2 axis bridge saw, the 3-axis double-knife saw and the middle block cutting bridge saw — plus a CNC stone router and a sawjet for the fifth position.
What axis count actually buys you

This is the part suppliers explain badly, and the part that decides your price.
Three axes — X, Y, Z. The blade stays vertical. It can move across the slab, along the slab, and down into it. That is enough for every straight cut, and for step cuts where two depths meet. It cannot produce a mitre, because the blade cannot tilt, and it cannot produce a curve, because the blade cannot rotate to follow the path.
Four axes plus one — the rotating head. Adding a rotating C-axis lets the blade turn in the horizontal plane. The practical result is that a mitre becomes a program step rather than a second setup on another machine. For a shop selling mitred worktops, this single axis replaces a machine and a handling operation.
Five axes plus two — rotation and tilt. Adding tilt lets the head approach the slab at an angle, and the two extra interpolating axes let the tool follow a path in space rather than a line in the plane. That is what makes sink cutouts with radiused corners, arcs, and shaped pieces possible without a waterjet.
The trade-off is not free. Every added axis is another thing to calibrate, another thing the operator must understand, and another source of error when a job is set up wrong. A 5+2 machine programmed badly produces worse parts than a 3-axis machine programmed well. If nobody on your floor will own the programming, buy the simpler machine and subcontract the shapes.
Five questions that decide the configuration
1. What slab sizes and weights do you actually handle?
Not the maximum you have ever seen. The distribution across your last fifty jobs. Table size and suction capacity are priced on the maximum, and you pay for it on every job.
2. Which cut types dominate your quotes?
Count them. If 80% of your value is straight dimensional work, a 3-axis machine with a good table earns more per euro than a 5+2 running the same program. If mitres appear on half your jobs, the rotating head pays for itself in handling time alone.
3. What throughput do you need per shift?
Throughput comes from setup time and cut speed together. A machine that needs forty minutes of setup per job loses to a slower machine that needs five. Ask what the control does to reuse a previous job — nesting memory, saved programs, and probe routines matter more than a headline feed rate.
4. What power, water and foundation do you have?
This is the question that most often invalidates a shortlist after the price has been agreed. Connected load, water supply and drainage, and floor loading all constrain which machines can physically be installed in your building.
5. Who programs it, and who maintains it?
Programming is a skill, not a feature. Ask who will run the control, what training is included, and how quickly a wear part arrives. A machine with no local support is a machine that stops earning the first time a bearing fails.
What the quotation must state
A quotation that lists only a model name and a price is not a specification. These are the fields that decide whether the machine matches your production, and every one of them should be in writing before you compare offers:
| Item | Why it matters |
|---|---|
| Maximum slab size and table dimensions | Decides which jobs you can load without re-cutting |
| Maximum cutting thickness | Decides whether you can cut thick countertop blanks |
| Blade diameter and flange size | Sets your maximum depth of cut and your blade cost |
| Spindle motor power | Sets what the machine actually cuts, not what it is rated to touch |
| Axis travel on X, Y, Z | The real working envelope |
| Rotating and tilting head range | Decides which mitre and bevel angles are possible |
| Control system and servo type | Decides programming effort and spare-part availability |
| Suction cup count and holding capacity | Decides whether polished slabs stay put |
| Water requirement and drainage | Decides installation cost you did not budget |
| Total connected load | Decides whether your supply can run it |
| Wear-part list and lead time | Decides how long a breakdown costs you |
| Warranty terms, training and installation scope | Decides your real first-year cost |
If a supplier cannot state a number, that absence is itself information about the supplier.
Three tests before you accept the machine
1. Cut your own material, not their sample. Bring the granite you actually sell — the hardest, the most veined, the one that chips. A machine demonstrated on soft material proves nothing about your production.
2. Measure the piece, do not look at it. Squareness across the diagonals of a full-size piece, and parallelism between two cut edges. A machine can look accurate and be out of square by enough to fail on installation.
3. Run a shift-length trial. Cut for several hours, then check three things: chipping at the exit edge, blade wear, and whether the tenth piece matches the first. Repeatability over a shift matters more than the first cut.
Running cost: the four lines that decide payback
Machines are compared on purchase price and paid for on running cost. There are four lines:
- Blades. The dominant consumable. Cost per metre cut depends on the granite, the blade specification and the operator. Ask for blade consumption data on material similar to yours.
- Water and power. Both are priced per hour of running, and both scale with how much idle time the machine has.
- Wear parts. Linear guides, seals, suction pads, bearings. These are predictable, which means they should be quoted, not discovered.
- Labour and skill. The expensive line is usually the operator standing next to a machine that is waiting for a program. Setup time is a labour cost multiplied by every job you run.
To estimate cost per cut without supplier data: take your blade cost and divide by the metres a blade realistically yields on your material, then add power, water and operator time per job. If the sum is close to your selling price per metre, the machine is the wrong machine regardless of what it cost to buy.
Design details worth paying for
The features that show up on working machines, and what each one is for:
- Milling cutter with edge grinding and shaping capability. Turns a cutting machine into a machine that can also dress and shape, which removes a second handling step for shaped work.
- Absolute servo motors with memory. Absolute encoders know their position without re-homing after a power interruption — the practical value is that an interrupted job can resume instead of being scrapped.
- Large-capacity suction cups with an independent worktable. Large cups hold polished slabs without point loading; an independent table lets you set up the next slab while the current one is being cut.
- Drilling cutter capability. A cutter that can be installed in the head lets one machine produce tap holes and small openings without a second setup.
- Cantilever operation panel. The control sits where the operator can see the cut line and the screen at the same time, which shortens every setup.
- Oil-immersed linear guides. Guides that run in oil wear more slowly and hold accuracy longer than guides that depend on grease intervals being remembered.
Mistakes that cost the most
- Buying axes you cannot program. The most expensive machine in a shop is the one whose capability nobody uses.
- Sizing for the largest slab instead of the common slab. You pay for table and suction capacity on every job you never run.
- Ignoring water, drainage and floor loading. These are installation costs that arrive after the price is fixed.
- No spare-part commitment. Ask for the wear-part list and the lead time in writing; a machine that stops for three weeks costs more than the price difference you negotiated.
- Expecting one machine to cut, mitre and do cutouts at production speed. It can usually do all three, slowly.
- Buying on feed rate. Setup time and repeatability decide your output, not the fastest number in the brochure.
FAQ
Is a 3-axis bridge saw enough for countertop production?
For straight dimensional work with square edges, yes — that is exactly what three axes do. It stops being enough when you sell mitred edges or internal cutouts, because those need a head that rotates or a tool that can travel inside the slab.
Can a bridge saw cut a sink opening?
Only if the head can be positioned to enter the slab away from the edge and interpolate a path. On a 3-axis machine that generally means a separate cutter and a drilling step; on a 5+2 machine it is a programmed operation.
Do I need a waterjet instead?
If your value comes from complex shapes, tight nesting and mixed materials, yes. If your value comes from volume on straight cuts, a waterjet will be slower per metre and more expensive to run. Many shops need both eventually, but rarely at the same time.
What slab size should the table take?
The largest slab in your actual order book, not the largest size available in your region. Oversizing the table raises the price and the floor space for capacity you will not use most weeks.
How long does it take an operator to learn the control?
It depends on the control and on whether the supplier trains on your own jobs. Ask for training that uses your cut list, not a demonstration file, and ask how a program is saved and recalled — that is what an operator uses every day.
New or used?
A used machine with a documented service history and available spare parts is a reasonable buy. A used machine with an obsolete control is not, because the control is the part you cannot source locally.
Getting a configuration you can actually use
To get a configuration that matches your production rather than a brochure, send three things: a sample of your cut list, your common slab sizes, and your required output per shift. Those three answers decide the machine family before any price is discussed.
If your next step also involves finishing, the edge polishing group is the other half of the same decision, and the buyer’s guide to stone edge polishing machines covers how to specify that machine. Blade specification is its own subject — the granite and marble saw blade range is where that decision starts, and the CNC bridge saw category lists the current models.
