When I went through MOZA's NexGen algorithm version numbers I came out the other side with one conclusion: the software badges are close to meaningless as a buying signal. The entry-level R5 Pro carries a higher NexGen number than the flagship R21 Ultra. That tells you nothing about how a base feels.

The hardware is a different story. There is exactly one real generational change in MOZA's current range, and it is mechanical: the Ultra bases use a flat-wire motor, the R-series V2 and V3 bases do not. That is the whole argument for paying Ultra money. So it's worth understanding properly — including the parts MOZA's own pages get wrong.

What a flat-wire motor actually is

Every direct drive base is built around a permanent-magnet motor. Copper wire is wound into slots in the stator, magnets sit on the rotor, and current through the copper produces torque. Standard practice for decades has been round wire — circular-section conductors, wound into the slots.

Flat wire uses rectangular-section conductors instead. Stack rectangles into a slot and you waste far less space than stacking circles. More copper in the same slot area means more torque from the same physical motor, and the copper sits in better thermal contact with the stator iron, so heat gets out faster.

This isn't exotic. Flat-wire (or hairpin) windings are standard in modern EV traction motors, precisely because packaging and thermal limits are what constrain those machines. MOZA's announcement frames it that way — "widely adopted in new-energy vehicles, aerospace engineering and more" — and claims the R21 Ultra and R25 Ultra were the world's first sim racing wheelbases to use it.

So the two genuine, physics-backed wins are torque density and thermal performance. Hold that thought, because MOZA leads with a third claim.

The cogging claim, and the part that's missing

MOZA's headline on every Ultra page is "All-new Zero-cogging Flat-wire Motor". Cogging is the notchiness you feel turning an unpowered direct drive wheel slowly by hand — the rim seems to step from one magnetic position to the next. It comes from the rotor magnets interacting with the stator's slot geometry, and it exists with no current flowing at all. On a wheelbase it's the enemy of small corrections around the straight-ahead position, and of feeling the moment a front tyre lets go.

MOZA does state a mechanism, in the R21/R25 announcement: "With reduced stator slot size in the flat-wire motor, cogging is eliminated."

Narrower slot openings genuinely do reduce cogging — that part is sound engineering. But notice what it means: the improvement comes from a slot geometry decision, not from the shape of the wire. Round-wire motors can and do use narrow slot openings, skewed rotors and shaped magnets to attack cogging, and that is precisely what MOZA already did on the R16 V2, whose page advertises an "asymmetric segmented skew design" that "significantly reduces cogging torque".

Flat wire's inherent contribution is torque density and heat. The cogging win is a design choice made at the same time. MOZA's marketing collapses two different things into one bullet.

There's also a physics ceiling: in any slotted motor cogging goes very low but never to zero. Genuinely zero-cogging machines use slotless or ironless topologies — see the Simucube motor approach for the contrast.

MOZA can't decide whether it's "zero" or "ultra-low"

This isn't me splitting hairs — MOZA contradicts itself on single pages.

The R21 Ultra page runs a headline reading "All-new Zero-cogging Flat-wire Motor". The icon row at the top of that same page reads "Ultra-low Cogging Torque". The R16 Ultra product description says the base "combines sustained 16 N·m torque, a next-generation high-density flat-wire motor, ultra-low cogging" — while the marketing block further down the page says "With true zero cogging".

"Ultra-low" is the honest engineering description and I'd bet it's what the spec sheet says internally. "True zero" is the one that made it into the headings. When a manufacturer publishes both on the same page, believe the conservative one.

What actually changed: same torque, much smaller box

Here's the evidence I find genuinely convincing, and it's buried in MOZA's spec tables rather than its headlines.

The R16 V2 measures 327 × 170 × 130 mm and weighs 8.9 kg. The R16 Ultra measures 206 × 140 × 135 mm. Same 16 N·m. That's about 46% less volume and 121 mm shorter, for identical claimed torque and an identical 21-bit encoder at 2,097,152 counts per revolution.

That is what a torque density improvement looks like from the outside, and in a tight rig or on a wheel stand it's the change you'll notice. Look at the R21 Ultra too — 248.8 × 140 × 135 mm, identical cross-section to the R16 Ultra, just longer. One motor platform with different stack lengths, exactly how you'd build a range around a new winding.

MOZA doesn't publish the Ultra bases' weight, which is a shame — weight is the number that would settle the torque-density argument outright.

The 12% versus 26% problem

MOZA quantifies the claim, and the quantification doesn't hold together. The R16 Ultra page says its flat-wire motor "offers 12% higher torque density than similarly sized competitors". The R21/R25 announcement says the same "next-generation custom flat-wire motor" offers "26% higher torque density than similarly sized competitors".

Same phrase, same motor family, one figure double the other. So "the flat-wire motor" is not one motor, and the percentage is a per-model result rather than a property of the technology. MOZA never names the competitors or the measurement method either — so the claim can't be checked in either direction. Treat both numbers as marketing and the dimensions as evidence.

Peak or sustained? MOZA's own store disagrees with MOZA's own FAQ

This matters more than the cogging semantics: it's the difference between a base that holds its rated force through a long stint and one that backs off.

MOZA's US wheel-base collection page badges the R16 Ultra as Peak Torque 16 Nm. The R21 Ultra and R25 Ultra on the same page are badged Sustained Torque. But the R16 Ultra's own product FAQ says the opposite: "The R16 Ultra delivers 16 N·m of sustained direct drive torque, maintaining consistent force feedback through demanding corners and extended racing sessions." And the R21/R25 announcement describes those two as "delivering 21-25Nm of peak torque" — the reverse of their own store badges.

So MOZA's peak-versus-sustained labelling is unreliable across the entire Ultra line. If sustained output is your deciding factor, ask MOZA support and get it in writing before you spend the money.

Both Ultra pages also headline "CNC-machined Aluminum Housing" and then call the part "a rigid die-cast aluminum unibody" — two different processes.

What the R16 V2 was already doing

Worth being fair to the outgoing base, because the upgrade case is narrower than it looks. The R16 V2 already ran an asymmetric segmented skew rotor, an 18-slot/16-pole layout to minimise torque ripple, a carbon-fibre-wrapped rotor and NexGen 4.0's cogging torque compensation in software. Its own page claims "practically zero torque ripple, notching or cogging". MOZA's R16 Ultra FAQ is refreshingly straight about it: "The upgrade is about refinement and control rather than simply increasing torque output."

What genuinely improved: the motor, the processor (280 MHz to 600 MHz), the packaging, the slip-ring rating (over 5 million revolutions to over 10 million), and native bottom, front and side mounting rather than an optional bracket. What didn't change: the encoder, or the headline torque figure.

So is it worth the money?

US prices, checked on MOZA's own store today: R9 V3 $299 (from $349), R12 V2 $399 (from $469), R16 Ultra $529, R21 Ultra $699, R25 Ultra $899. The Canadian storefront has the R16 Ultra at CAD $729 with dispatch estimated late August 2026, so stock is only just arriving as I write this.

My honest read, and it's consistent with what I said in the MOZA buying guide:

  • Already own an R16 V2 or R12 V2? Don't upgrade for the motor. You won't find 12% torque density on a lap chart, and your V2 already fights cogging three ways. Spend it on pedals.
  • Buying new around $500? The R16 Ultra at $529 against the R12 V2 at $399 is a genuinely good $130 — four extra newton-metres, the newer motor, the faster processor, and a much smaller box. That's the one clear recommendation in this range.
  • Chasing the flagship? The R21 Ultra's plastic rear cover on a $699 base is worth knowing about, and the jump from $529 to $699 buys 5 N·m you'll probably dial back anyway in Pit House.

The point I keep coming back to on why direct drive costs what it does: the motor is where the money goes, so the motor is the spec worth reading — the spec table, not the headline. For the cross-brand view at these prices, the four-brand comparison puts these numbers against Fanatec, Logitech and Thrustmaster.

FAQ

Does a flat-wire motor really have zero cogging?

No. It has very low cogging. MOZA's own pages say "ultra-low cogging torque" in one place and "true zero cogging" in another, on the same product. A slotted permanent-magnet motor always has some cogging; only slotless or ironless designs get to genuinely zero.

Which MOZA bases have the flat-wire motor?

The Ultra line: R16 Ultra, R21 Ultra and R25 Ultra. The R9 V3, R12 V2 and R16 V2 use MOZA's earlier round-wire servo motors with a skewed segmented rotor instead.

Is the R16 Ultra's 16 Nm peak or sustained?

MOZA publishes both answers. The US store badges it "Peak Torque 16 Nm"; the product FAQ on the same site says "16 N·m of sustained direct drive torque". Get clarification from MOZA support before buying if sustained output is your deciding factor.

Is it worth upgrading from an R16 V2 to an R16 Ultra?

For the feel, no — the V2 already targets cogging with a skewed rotor, an 18-slot/16-pole layout and software compensation. For packaging, maybe: the Ultra is roughly 46% smaller by volume for the same torque, which matters in a cramped rig.

Does the flat-wire motor change my FFB settings?

Not fundamentally. The tuning logic is the same as any MOZA base — the FFB curve tool and the settings approach I use on the R9 both transfer. Lower cogging means you can usually run less artificial damping and friction before the wheel feels sticky.

Is the R16 Ultra Xbox or PlayStation compatible?

PC natively. Xbox only when paired with the officially licensed MOZA ESX steering wheel — the base alone does not provide Xbox compatibility. PlayStation is not officially supported. If you need console support, the direct drive explainer covers which brands actually hold the licences.