Mazda’s ‘G-Vectoring Control’ tech review

Mazda has new tech to make life easier and smoother behind the wheel

Mazda 6 G Vectoring

Mazda has patented a forthcoming technology it calls ‘G-Vectoring Control’, set to debut in production form on the facelifted Mazda 3 range next month but tested here on a re-jigged Mazda 6.

A purely software-driven device that aims to enhance on-centre steering connection and turn-in response. Eight years in the making, it’s essentially an automation of the techniques that expert drivers use to subtly transfer some weight onto a car’s nose to improve its turn-in to a corner and tighten its line.

Because G-Vectoring Control, or GVC, is about to be rolled out across Mazda’s entire range, starting with August’s facelifted Mazda 3 range (but missing July’s all-new CX-9 SUV, which will have to wait until update time). However, Mazda admits that adding GVC to its existing line-up may occur during a model-year update, not merely as an accompaniment to a mid-life facelift, so there’s a strong chance the MY17 Mazda 2 and 6 will be next, with Mazda CX-3 and CX-5 to follow.

Mazda’s G-Vectoring Control

That’s a difficult question to answer, given that G-Vectoring Control is more of an engineering philosophy than an actual model. But if it achieves what Mazda expects it to, we may well find ‘GVC’ in a whole bunch of competitor products in the future, with royalties heading to Hiroshima.

If you remember how Radial Tuned Suspension (RTS) transformed Holden’s line-up in the late-’70s, there’s a chance GVC will achieve an element of that with Mazda’s current line-up – not only for the effective, if subtle, improvement it brings to on-centre steering smoothness/response and turn-in alacrity, but because it will debut in unison with completely refreshed suspension tunes (that will hopefully realise the dynamic potential of Mazda’s SkyActiv range).

PLUS: Subtly improved on initial acquaintance, but of greater worth the further and faster you drive it …
MINUS: Mazda admits that not everyone can appreciate GVC’s virtues, or even feel the difference, journalists included…


REMEMBER the Energy Polariser, a collection of crystals in an engine-bay box that apparently did a bunch of esoteric stuff to Brocky’s HDT VL SS Group A and VL Director back in 1986/87, without the actual hard evidence to prove it?

Well, you could be forgiven for thinking that Mazda’s G-Vectoring Control (or GVC) – a patented technology about to be progressively rolled out across its entire ‘SkyActiv’ passenger-car range – somehow fits into the same category. But this time the Japanese have the data to back up their intriguing, almost subversively effective approach to improving dynamic flow, or what Mazda calls “the coordination of motion”.

The aim is simple: enhance the connection between vehicle and driver, thereby reducing the level of guesswork when guiding a car around the straight-ahead position and through the first quarter-turn or so of lock, and into a corner, particularly with a constant throttle position.

Any mistake on turn-in will resonate through an entire corner, making that first movement arguably the most crucial aspect of how a car strings itself together dynamically. And one that hasn’t exactly been a Mazda highlight in recent years, particularly our Euro-tuned models that are less lively around straight ahead in order to feel stable at autobahn speeds.

Mazda’s G-Vectoring Control

But G-Vectoring Control is intended to solve all that. By retarding the ignition slightly to decrease engine torque enough to transfer some weight onto the front end of the car – a seemingly insignificant 5kg – the electronics replicate what a good driver would do when lifting or feathering the throttle, or by left-foot braking (while still on the throttle) when entering a corner.

Thing is, this happens in just 50 milliseconds – faster than the human body can detect – so part of the reason it has taken until now for G-Vectoring Control to surface is that it needed an engine responsive and linear enough, as well as an ECU with the processing power, to enable the adjustments to happen so quickly.

The intended outcome of G-Vectoring Control’s effectiveness is not only improved feedback and greater fluidity of movement but also comfort. As Dave Coleman, a Mazda North America chassis engineer, pointed out: “if you make a car really direct, really intuitive, really linear, it’s better for everybody in the car”.

Mazda’s G-Vectoring Control

And so the exercises begin. The vehicle involved is a current 2016 Mazda 6 Atenza auto petrol sedan, re-jigged with a dashboard button to allow GVC to be turned on and off, enabling a direct comparison.

Traversing a course outlined by witch’s hats, with a straight on either side and slightly jagged corners on either end – one of which is wet – we set the 6’s cruise control to 18mph (29km/h) and start lapping. Sticking tight to the orange hats, we perform two clockwise laps with GVC turned off, and another two with it switched on. And then repeat to process, swapping between each mode with greater frequency.

Guiding the 6’s lovely three-spoke steering wheel with my fingertips, the difference is immediately obvious. Subtle, yes, but also clearly evident. Where the standard mode transmits little true connection through the electric steering at straight ahead and through the first phase of turn-in, GVC sparks a connection, enabling easier and sweeter placement of the 6’s nose into each corner, and less understeer through the wet corner.

Mazda’s G-Vectoring Control

Given the fact there are no hardware changes, how can GVC possibly reduce understeer? Well, simply by removing the guesswork. Without any real feedback until you’re already in the corner – a common criticism by Wheels of Mazda’s current SkyActiv product – the driver tends to apply more steering lock than required entering the wet corner, which promotes understeer, even though the difference between each arc (GVC on and off) isn’t huge.

Another benefit of improving the steering intuitiveness via GVC is greater passenger comfort. When faced with the relative dead zone of the standard car, the driver unintentionally adds more lock than required and does so with greater speed. The result rocks passengers around noticeably more than with GVC operating, simply because the driver’s inputs are crisper and more measured when there’s a tangible connection between the tyres and the wheel rim.

We perform a similar exercise on gravel, where the difference in understeer between the standard set-up and GVC is even more marked, albeit hardly chalk and cheese. But without GVC doing its business, the 6’s driving line is messier, no matter how precisely I try and drive it. Simply put, there’s more guesswork in guiding the standard 6 with half a turn or so of steering lock on than there is when GVC is assisting the process.

Mazda’s G-Vectoring Control

To further demonstrate the reduction in head toss when G-Vectoring Control is smoothing the transition between the initial phase of diagonal roll and subsequent turn-in, we perform lane-change exercises on Laguna Seca Raceway’s main straight. And it’s clear there’s a definite sweetening in the car’s fluidity of movement, though again, you need to be paying attention to really notice it.

Finally, we point the GVC-equipped 6 out onto the main road and up into some hills not far from the racetrack. Again, it’s a cruise-controlled exercise – 45mph (72km/h), which is the speed limit on this section of Californian road – and while the difference between each mode is arguably less transparent through these constant-radius corners than it is in other situations, there’s still a subtle improvement in the ease of placement into and through a corner.

What I wasn’t expecting, however, was to notice a difference on faster, relatively straight roads. Cruising at an indicated 70mph (112km/h) or so, the deadness of the standard electric steering, and the driver’s ability to ‘chase’ the steering either side of straight ahead, somehow vanishes when GVC is operating. Wheel movements become smoother because you feel more connected to the car, and this relaxation in driver input subtly benefits everybody because the car seems to track straighter.

Mazda’s G-Vectoring Control

There will definitely be sceptics. Some colleagues claimed they couldn’t tell the difference, and even some of Mazda’s own employees felt no discernible change in the GVC-equipped 6s they tested in the real world. But the Japanese have the data to prove it – smoother driver inputs, less wheel movement, and the resulting tighter cornering line on slippery surfaces and less passenger body movement and head toss.

It should be pointed out that the US-spec 6 we drove was deemed the only current SkyActiv Mazda vehicle whose chassis set-up was good enough to work effectively with G-Vectoring Control.

When GVC begins rolling out across Mazda’s global SkyActiv range next month (starting in Australia with the facelifted 3), it will be accompanied by a comprehensive chassis re-tune. And the system will be standard across the board, without the switchable arrangement of the pre-production cars we drove.

Hopefully, given G-Vectoring Control’s subtlety of operation and the latent dynamic potential in Mazda’s SkyActiv products, we’ll truly get to experience “best-in-class driving dynamics”. Come late-July and the updated 3, we’ll know for sure.

Model: Mazda 6 Atenza ‘GVC’
Engine: 2488cc 4cyl, dohc, 16v
Max power: 138kW @ 5700rpm
Max torque: 250Nm @ 3250rpm
Transmission: 6-speed automatic
Weight: 1501kg
0-100km/h: 7.6sec (tested)
Economy: 6.6L/100km
Price: $45,390
On sale: 2017


How are you finding our new site design? Tell us in the comments below or send us your thoughts at


Subscribe to Wheels magazine

Subscribe to Wheels Magazine and save up to 44%
Get your monthly fix of news, reviews and stories on the greatest cars and minds in the automotive world.



Nathan Ponchard

We recommend


Mk8 Golf

Volkswagen Golf Mk8 pricing and specs revealed

Revised and refreshed Golf hits the market to take on Asian rivals

2 days ago
Jordan Mulach
Please enable JavaScript to view the comments powered by Disqus.