The initial targets were 500 hp at a 7,500-rpm power peak and 410 lb.-ft. (556 Nm) at 4,750 rpm. “Then, pushing and stretching ourselves, we looked at 100 hp per liter and revised the target to that. Then we overachieved and delivered 526 hp at 7,500 rpm and 429 lb.-ft. (582 Nm) at 4,750 rpm. Our original target for engine speed was 8,000 rpm, then once we built hardware and tested and developed it, we verified that we could take that further out to the 8,250-rpm redline. And the power is pretty flat. It peaks at 7,500, and then drops off very slowly.”

The need for high torque in a wide rpm range led to big, long intake runners and the flat-plane crank. “We try to let attribute targets drive the hardware content required to deliver it,” Ladner says. “We’re always looking at what to do next to make the next step in performance, and the flat-plane crank was on the table as a potential technology.”

Achieving valvetrain stability at such high engine speeds presented a challenge, but meeting fuel-efficiency targets and emissions requirements did not. “We made all the right assumptions from the beginning,” he says, “so no major challenges there.”

But the flat-plane crank does bring an NVH challenge: “Lateral shaking forces are higher than we generally deal with on cross-plane engines,” Ladner says. “But while the order content and direction of vibration are different, it is part of our normal design process to look at vibrations for the engine system as well as for each individual component. We took into account the higher vibration levels as inputs to our CAE tools and [NVH] testing, and our design verification plans also involved a number of hours of shaker-rig testing. We just went through the design process as we normally would for any engine, but at higher-level inputs.”

While the engine team benchmarked the Ferrari 458 engine for performance, the vehicle team analyzed a front-engine Ferrari California for NVH. “We took measurements and looked at the parts and how they implemented some things,” Ladner says. “I think we beat it.”

It was not a cost-is-no-object program, but there was more freedom because the car would be priced accordingly.

“Again, lots of upfront planning,” Ladner says. “We put together assumptions of which parts would have to change, and it was pretty much everything. We have some history on what those types of parts cost in low volume, then put together a good cost projection and program target. It was a much more generous target than most volume mass-production programs, but managing and meeting that was maybe the biggest challenge.”

Still another hurdle, given the very low volume, was getting the right suppliers on board. “At these volumes it sometimes makes sense to go to special suppliers for certain things that may not be handled as well by bigger ones. In some cases, we have to be a little more creative to find someone we think will fit the part and the low-volume program.”

The engine is assembled by two skilled builders who follow it along a low-volume “niche” line in a separate building behind Ford’s Romeo engine plant, and they sign it when they’re done. Ford built just 100 GT350 and 37 (still more track-capable) GT350 Rs for 2015, the 50th anniversary of the original Shelby Mustang GT350, because that original 1965 run was 37. But for 2016, the production of regular models is expected to total 3,000-5,000 units.