What is the Difference Between CNC and MIM?
Last quarter a medical device company sent us a part they'd been machining for three years. Stainless steel bracket, 47 features, $23 per piece, 60,000 units annually. They wanted to know if MIM could cut their cost.
We ran the analysis. MIM tooling would cost $52,000. Per-piece cost would drop to $4.20. Break-even at 2,766 units, so they'd recover tooling investment in the first month of production and save $1.1 million annually after that.
They didn't switch.
The bracket had two bores requiring ±0.02mm for bearing press-fit. MIM at ±0.3% on those 18mm features would give ±0.054mm, nearly three times too loose. We could MIM the body and CNC-finish the bores, but that hybrid approach priced out at $7.80/piece. Still saves money, but not the dramatic reduction they wanted.
Tolerance
This is where most MIM projects die. MIM tolerance scales as percentage, ±0.3% standard, ±0.1% optimized. Engineers see ±0.3% and think "tight." On small features, it is.
| Feature | MIM ±0.3% | CNC |
|---|---|---|
| 10mm | ±0.03mm | ±0.025mm |
| 25mm | ±0.075mm | ±0.025mm |
| 50mm | ±0.15mm | ±0.025mm |
CNC holds absolute tolerance. MIM scales with size. That 50mm bore at ±0.05mm? MIM can't touch it without secondary machining.
But when tolerances work out, the savings are real. Spine Wave's surgical instrument program is the case we reference most because the numbers are public and the story is complete. They came to their MIM supplier with a shipping lock component for the Velocity® spinal fusion device. The part was machining at $18/piece, not outrageous for a medical component, but they needed 75,000 units per year. Annual machining spend: $1.35 million.
The MIM evaluation took four months. Tooling cost $38,000. First articles required two sintering cycle adjustments to hit dimensional specs. The per-piece cost landed at $2.80. First year total: $248,000 including tooling amortization. They saved over a million dollars in year one. By year three, cumulative savings exceeded $3.2 million.
Their quality engineer's comment after 18 months of production: "For a quality engineer, having nothing to report is a resounding endorsement."
No field failures. No dimensional drift. No supplier quality issues. The full case study is published on optimim.com. 85% cost reduction. Not every program hits those numbers, but 50-70% savings on complex stainless geometries above 20,000 annual volume is what we quote regularly.
Process
CNC machining is straightforward. Bar stock gets fixtured in the machine. Cutting tools follow programmed paths and remove material until your geometry remains. Change the program, get different parts. Lead time to first articles: days. Design changes cost programming time, not tooling.
MIM is a sequence of transformations. Metal powder, 2 to 20 microns in diameter, gets blended with thermoplastic binder at roughly 60% powder to 40% binder by volume. The mixture behaves like plastic during injection. It flows into mold cavities and solidifies into "green parts" that look correct but have no structural integrity.
The green part goes through debinding, thermal decomposition or solvent extraction, to remove the polymer. What remains is a "brown part": metal powder held together by friction and residual binder. Fragile enough to crumble if you handle it wrong.
The brown part enters a sintering furnace running at 85-95% of the alloy's melting point. Over 15-20 hours, solid-state diffusion bonds the powder particles. The part shrinks 15-20% in all dimensions as porosity collapses and density increases. A part molded at 60mm comes out of sintering at 48-51mm.
The shrinkage is predictable in aggregate but not perfectly uniform. Gravity affects parts differently depending on furnace orientation. Mass distribution influences local shrinkage rates. Wall thickness variation creates differential shrinkage that warps parts.
We've seen parts come out of sintering dimensionally correct and parts from the same mold come out twisted because someone loaded the furnace tray wrong. That kind of thing happens more than suppliers like to admit. Sintering fixture development, supports that hold part geometry during the hours-long thermal cycle, is engineering work, not a commodity. Suppliers without experience on geometrically similar parts will iterate through fixture designs at your expense.
The Failures
At least twice a month someone sends us an aluminum part asking for MIM pricing. The answer is always no. Aluminum powder forms an oxide layer that won't break down during sintering, the particles can't bond. Research labs have demonstrated aluminum MIM in controlled conditions. Commercial production doesn't exist. Aluminum means CNC.
Atmosphere control in the sintering furnace determines whether your material actually performs to spec. Stainless steels need hydrogen or vacuum atmosphere. Wrong atmosphere, or atmosphere contamination from a dirty furnace, causes chromium to migrate away from the surface. The parts look fine. Dimensions check out. But corrosion resistance is destroyed because surface chromium is depleted. We've rejected incoming MIM parts from other suppliers that passed dimensional inspection but failed salt spray testing in under 100 hours.
MIM parts retain 1-4% porosity. Those voids trap whatever contamination exists in your environment: machining coolant, cutting oil, atmospheric moisture.
We had a customer two years ago who specified MIM on structural brackets for cost reduction. Production ran smoothly for eight months. Then their field service team tried to weld a bracket during equipment repair.
The weld failed. Repeatedly.
Trapped contamination vaporizing through the weld pool created porosity and inclusions that destroyed joint integrity. They had to qualify a new CNC supplier for replacement parts. The "cost savings" from MIM turned into emergency procurement costs, field service delays, and a customer relationship problem.
If your parts might ever need welding, for assembly, repair, or field modification, MIM is disqualified.
Fatigue is another thing we get questions about. Wrought bar stock machines at 100% density with consistent grain structure. MIM sinters to 96-99% density with distributed microporosity. Static strength is equivalent. We've tested 17-4PH MIM parts against wrought specs and they match on tensile, yield, and hardness. 17-4PH H900 runs 1070-1190 MPa UTS, 970-1090 MPa yield, 35-40 HRC. 316L hits 520 MPa minimum with excellent corrosion resistance when sintering atmosphere is properly controlled. But those micropores act as stress concentrators under cyclic loading. For parts seeing 10^7+ cycles, rotating shafts, vibration-loaded brackets, anything with resonance concerns, wrought material outperforms MIM. Hot Isostatic Pressing can close porosity and recover fatigue performance, but that adds $10-25 per part.
Surface finish: MIM as-sintered achieves Ra 0.8-1.6μm, smoother than standard CNC at Ra 3.2μm. Density runs 96-99%, or 99%+ with HIP. Standards if you need them: MPIF 35-MIM (2025), ASTM B883-19, ISO 22068:2012.
Volume
| Annual Volume | Recommendation |
|---|---|
| Under 1,000 | CNC, tooling investment makes no sense |
| 1,000-5,000 | Usually CNC, run break-even math on complex parts |
| 5,000-20,000 | Depends on geometry, worth calculating |
| Above 20,000 | MIM usually wins if tolerances work |
Break-even formula: MIM tooling cost divided by per-piece savings. A $40,000 mold with $14/piece savings breaks even at 2,857 units. Above that volume, every part saves $14. At 50,000 annual volume, that's $700,000 per year.
Only about half of MIM molds we quote ever go into production. Design changes after tooling starts cost $2,000-15,000 per modification. Programs that chase cost reduction before design stability end up paying for tooling twice.
CNC machines anything you can fixture. Aluminum, titanium, Inconel, brass, plastics, composites, no restrictions beyond machinability. MIM material selection depends on established feedstock formulations. Stainless steels make up 50-57% of global MIM production. 316L for corrosion resistance, 17-4PH for strength, 420 for wear. Low alloy steels (4140, 4340, Fe-2Ni) for heat treatment or magnetic properties. Tool steels M2 and H13 for wear resistance, though carbon control during sintering requires careful process development. Titanium Ti-6Al-4V for medical and aerospace, sintered under vacuum. What doesn't work: aluminum, copper alloys, pure metals.
Material utilization: CNC converts 20-50% of your billet into parts, the rest becomes chips. Complex parts might waste 70-80%. MIM feedstock utilization runs 95-97% with runner recycling. For titanium at $60/kg, a part machined from 180g billet to 40g finish wastes 140g of material, $8.40 in scrap before you cut a single chip. MIM produces that part from roughly 42g of feedstock.
Our standard advice: prototype in CNC regardless of production intent. Validate fit, function, design intent. Lock the geometry. Then invest in MIM tooling if volume supports it.
When It Works
MIM wins when you have complex geometry requiring multiple CNC operations. Internal features, undercuts, thin walls, features that would need 5-axis work or EDM. Parts where machining cost runs $25-50/piece often quote under $5/piece in MIM.
AFT's turbocharger vane program: 180+ million parts produced for BMW, Ford, Tesla, Caterpillar. Airfoil profiles holding ±0.015mm, tighter than standard MIM through optimized tooling and process control. 20% cost reduction versus precision casting at automotive volumes. Documented on pim-international.com. That program required years of development investment, not something you replicate on a six-month timeline.
CNC wins on aluminum and copper parts, tight-tolerance features below ±0.1%, applications requiring weldability, programs needing full material traceability (aerospace, nuclear), parts under 5,000 annual volume, and geometries that machine efficiently. Simple turned parts on CNC lathes may never justify MIM tooling regardless of volume.
Hybrid works when you need MIM economics on complex geometry but have features requiring CNC precision. MIM the body, machine the critical bores.
Send us your part with annual volume. We'll quote both processes and run the break-even analysis. We'll tell you which features cause MIM problems and whether hybrid manufacturing makes sense. We don't push process recommendations based on what makes us more money. We push them based on thirty years of knowing which projects succeed and which ones turn into quality escapes and customer problems. A failed MIM program costs us more in engineering time and relationship damage than we'd ever make on the tooling. If your part fits MIM and your volume supports it, we'll show you numbers that look like the Spine Wave case, 70-85% cost reduction, reliable production, quality that doesn't require constant attention. If it doesn't fit, we'll tell you that too.