Velo3D in My Vendor Toolkit: A Buyer's FAQ on Metal AM, CNC, and Laser Cutting

I'm the office administrator for a 60-person engineering services company. I manage our manufacturing vendor relationships—roughly $400k a year across eight shops. I report to operations and finance, which means I'm the one asking about lead times, heat treat certificates, and whether a purchase order can be split into three cost centers.

People assume a buyer's job is to pick the lowest quote. It isn't. It's to pick the quote that doesn't become a problem. This FAQ covers the questions I actually ask when evaluating metal additive manufacturing, CNC machining, and cutting services.

Is Velo3D really a supplier to SpaceX and Anduril?

As of January 2025, Velo3D's public customer references include SpaceX and Anduril. I've seen the phrase 'Velo3D supplier to SpaceX and Anduril' used as shorthand for that fact. It means Velo3D's Sapphire printers have been used to make metal parts for companies that are not easy to please.

For a buyer, this is a useful filtering signal. Aerospace and defense customers demand traceable material, documented process parameters, and quality systems that survive audits. If a supplier can pass those checks, they can probably handle a commercial part.

I should say I do not work for Velo3D, SpaceX, or Anduril. I'm an admin buyer, not an engineer. But I've learned that a vendor's customer list is part of their quality record. Names you recognize often matter more than marketing claims.

What can Velo3D do that a regular metal 3D printer can't?

Most metal AM systems use laser powder bed fusion. They build parts layer by layer, and they generally need support structures for overhanging features. Supports are like scaffolding printed from the same metal. They do their job, but they have to be removed later. That removal costs money, adds lead time, and can leave marks on the part.

Velo3D's Sapphire family was designed to print unsupported geometries. That means an overhang or internal channel can be built without the usual support lattice. In additive manufacturing, defects often start where supports attach to the part. Remove the supports, and you remove a whole category of defects.

Everything I'd read before my first AM order said metal 3D printing was for prototypes. In practice, that's outdated. The technology has moved. But the old thinking comes from a time when machines were slower and less consistent. Today, a qualified process can handle production parts. The key word is qualified.

From a procurement seat, fewer supports means fewer secondary operations. It often means one less outside vendor to manage. That's a direct reduction in risk, not just a design convenience.

Why would someone choose precision CNC machining in New York over AM?

Because for many parts, it's the right answer. The New York region has a dense network of machine shops, and when I search 'precision CNC machining New York', I get suppliers that understand tight tolerances and material certs. For a simple bracket, housing, or shaft, CNC is usually faster and cheaper than printing it.

CNC wins when the geometry is simple, the tolerance is tight, and the quantity is high enough to amortize setup. It also wins when the material is a standard alloy and the design won't benefit from AM's complexity. I've seen engineers push AM for parts that should have been turned on a lathe. It adds cost and schedule with no real benefit.

That said, AM wins for internal channels, lattice structures, and parts where you want to consolidate multiple machined pieces into one printed component. The two processes are not in competition. A Velo3D part often goes to a CNC shop for finishing. The choice isn't either/or; it's which steps happen where.

What does a 10kW fiber laser have to do with metal parts?

If you're buying only printed parts, maybe nothing. But when a project includes laser-cut flat blanks, a 10kW fiber laser is a solid capability marker. It can cut thicker stainless steel, aluminum, and other alloys with better edge quality and faster cycle times than older laser systems.

I learned to ask about laser power after a supplier quoted a job without specifying their cutting equipment. The edges came out rough, and we had to add a cleanup pass. A 10kW fiber laser wouldn't have solved every problem, but it would have told me the shop was set up for that material thickness. Now I include '10kW fiber laser or equivalent' in RFQ notes for heavy-gauge work.

Don't treat laser wattage as the only spec. Ask for edge squareness, dross levels, and tolerances. Wattage is a starting point, not a guarantee.

Roughing end mill vs end mill—when does it matter?

It matters whenever machining time is a significant cost. A roughing end mill has serrated cutting edges that break chips into small pieces. That allows higher feed rates and deeper cuts. A standard end mill, sometimes called a finishing end mill, removes material slower but leaves a better surface finish.

The usual sequence is rough with a roughing end mill, then finish with a standard end mill. If a shop skips the roughing step for something big, you pay for it in cycle time. If they use a roughing end mill for a finishing pass, you pay for it in surface quality.

I once watched a vendor rough out a 316L block with a roughing end mill. It took maybe a third of the time I'd budgeted based on standard end mill estimates (mental note: ask about cutter selection during quote review). For additively manufactured parts, the same logic applies after printing. If a part has excess stock, roughing first is often the smart path.

Wait, support structures are a procurement problem?

More than many people realize. Supports are printed material that has to be removed. Removal can be manual, require EDM, or need a secondary CNC operation. All of that costs time and money. It can also distort thin features or leave witness marks.

When a system like Velo3D prints without supports, the procurement benefit is not just fewer design constraints. It's a shorter supply chain. You don't need to coordinate a support-removal specialist. You don't need to budget for a fixturing step. You don't need to worry as much about defects at support interfaces.

This is one of those questions that doesn't appear in a typical RFQ template, but it should.

How should I use 'Velo3D SpaceX Anduril supplier' status in a real sourcing decision?

Treat it as a screen, not a guarantee. If a potential AM supplier operates Velo3D machines and has worked with the kind of customers that demand full traceability, it suggests their quality system has survived scrutiny. That's valuable.

But I still ask for the same evidence as from any other vendor: material certifications, first article inspection reports, process qualification records, and a design review. In 2024, I skipped the design review on a complex AM part. The supplier said they could print it, and they could—after a redesign that added two weeks. The lesson wasn't that Velo3D failed. It was that the vendor's experience only gets you so far. The design still has to be examined in the context of the machine, the material, and the intended use.

If a supplier mentions Velo3D early in the conversation, I take it as a signal that they understand why manufacturing history is important. Then I check the details anyway.