I recently saw an add for a 1978 Camaro with a 5.7 litre engine. I was surprised to see it only produces 170 hp. Similarly sized naturally aspirated engines today are closer to 450 hp. What are the main innovations in modern naturally aspirated engines that account for this difference?
Incidentally, I asked this question to chat gpt and got a bunch of dubious reasons including improved aerodynamics 
Catalytic converters were required starting in 1975, which meant automakers couldn’t use leaded fuel. This was a low point for engine performance, as most of the efforts to meet emissions standards involved choking off the airflow and lowering the compression. You could make power in 1978, just not cleanly.
The primary change following this era that allowed performance to creep back up was computerization.
An ICE is, funnily enough, a heat engine. So the efficiency is governed by the fundamentals of heat engines. So trying to optimise that is a good start. Best way of getting the thing to generate more power is to sweep more volume. If we have a fixed capacity engine, rev higher. Hard to beat that. An engine that won’t shake itself apart is a terrific start.
You can look to minimising losses in the engine. Of which there are lots. And funnily enough, for a hard revving engine, aerodynamic losses in the crankcase may matter. In the margins, but if you really care, it can help. Better production tolerances, higher quality oils, Lots of little gains. A surprising amount of a car engine’s power gets sucked up just keeping it turning.
Critically you look at combustion efficiency, and try to get at the energy in your fuel, although that isn’t a direct input into just total power. Very powerful engines may run particularly inefficiently. But for road cars, combustion efficiency has gone hand in hand with improved thermodynamic efficiency.
The fundamental thing with thermal efficiency is that that hotter your burn, the more energy you can get out of the combustion process. So a huge amount of work has gone into optimising fuel burn. This has been in concert with efforts to reduce air pollution and improve fuel efficiency. A modern high compression petrol engine with direct injection and stratified ignition is generations ahead of a carburettor with a coil and distributor, vacuum advance, and tuning done by ear with a screwdriver. The burn can be controlled with tiny nuances across the entire rev range. The fuel will almost all burn, and do so at as high a temperature as materials science will let it. Everything is a win. High performance electronics, especially engine management, that are able dynamically take into consideration the whole picture of the engine operations is a key enabler. Indeed for a long time the pacing item in being able to achieve these gains.
So four classes of optimisation:
Revs, thermal efficiency, combustion efficiency, losses.
As noted above, 1978 was probably the low in apparent engine performance. But a change to how power was measured and reported a few years earlier clipped at least a third off the reported numbers. But manufacturers really struggled to make their old clunkers work.
Mechanical design improvements that reduce friction inside the engine, higher compression ratios (which goes hand-in-hand with better materials and fabrication methods), computer-controlled variable fuel and air valve timing, improved airflow on the intake side, and improvements in exhaust flow.
I have always been amazed that the basic design for ICEs in vehicles - four stroke, cylindrical pistons, common crank shaft, etc. - hasn’t changed in eons. You’d think something revolutionary would come along after all these decades.* The only one I can think of is the Wankel, but it didn’t last long. Based on this, can it be assumed the design is optimized?
*Well, there’s electric of course. But that’s not really a new idea, either.
Mazda resurrects the Wankel engine as a hybrid:
The obscenely giant (and thirsty) 8 cylinder engines with rather low horsepower were almost exclusively an American phenomenon at that time. European and Japanese auto builders got much more horsepower out of much smaller engines even then.
IMHO it all comes down to, “We’re really good at making round pegs that fit nicely into round holes.” Most of the innovations in the basic layout involve straying from the round peg/round hole layout, which, as we saw with the Wankel, isn’t good for reliability or longevity. You just can’t get the rotor to seal well inside the housing, and so rotaries struggle with longevity and efficiency.
There’s plenty that’s not ideal with reciprocating pistons, and most of the efficiency improvements have been focused on the intake stroke – getting enough air into the engine while the pistons accelerate from a dead stop, decelerate, stop, and reverse direction 10 to 15 times a second. And so we went from flat heads to overhead valves, to overhead cams. We have variable ignition timing and variable valve timing, variable intake runners, etc. The compression stroke is equally limiting, and so we went from carbs to port injection to direct injection. We have domed pistons and hemispherical heads. 4 or 5 valves per cylinder with different sizes for individual intake valves. More recent innovations have involved pre-combustion chambers and the like.
In any case, what’s really impressive isn’t that we can get 450+ hp out of a 5.7L V8 now, it’s that we can do it while getting 34mpg highway and not spewing a bunch of particulates. People have been complaining about the death of cars for a long time, but we’ve been in a golden age of automobiles since computer engine management revolutionized things through the late 80s and 90s, and the golden age just keeps on glistening.
Inventors and companies have been trying to improve upon the reciprocating engine essentially since it was invented, including dual action or “opposed” piston, radial piston arrangements, and rotating ‘pistons’ like the Wankel rotary piston that directly produce rotational power without a crankshaft. However, the single-ended piston engine in both spark- and compression-ignition has some unique advantages of being compact, relatively mechanically simple (despite the crank arrangement), thermodynamically straightforward, and most importantly can deliver near optimal work across a wide powerband. By comparison, dual action pistions like the Stirling cycle are mechanically complex and difficult to scale up because of packaging, rotary engines are compact and deliver high power but are difficult to get complete combustion and suffer from mechanical wear at the seal interfaces, and turbines can offer high specific power and fuel efficiency but only across a narrow powerband and have to run at really high speeds requiring frequent bearing maintenance and producing annoying noise and vibration characteristics.
Although the Stirling cycle is basically the jewel of thermodynamic efficiency for a reciprocating engine, the Otto and Diesel cycles (and modifications thereof like the Atkinson cycle) can be well optimized and conveniently packaged, and despite the claim that electrification will dispense with the ‘inefficiency’ of reciprocating heat engines, the latter don’t seem likely to disappear any time soon even if they are rapidly being eliminated in the small engine category of applications like lawn/garden maintenance or light vehicles because of their specific power, convenience, and flexibility. And despite someone coming along every few years with some supposed genius innovation with a six-stroke cycle or epicyclic action there just really isn’t much design space left to innovate without adding an opprobrious degree of complexity. You can build a simple single action piston engine running on basically any volitizable fuel with just a few parts, and attach it to a crank to get rotary motion, making it essentially a combination of three simple machines, and even if you only get less than 25% thermodynamic efficiency it still puts enough power in a person’s hand to cut down a try or traverse long distances in a fraction of the time it would take with human or animal labor.
Stranger
I’ve seen similar improvements. My first car was an '82 Cutlass Supreme:
My current car is a 2016 Infiniti Q50:
Power output is the product of torque and RPM; if you can make the engine spin faster and produce more torque at the same time, your power output goes way up . If you divide torque by engine displacement, you get a sense of how fully the engine’s displacement is being utilized. On a per-volume basis, my current car makes 44% more torque than my first car. So let’s start with how they do that:
Better breathing. If you want to burn fuel, you need air. Four valves per cylinder means two intake valves instead of one, meaning more flow area for mixture to flow into the cylinder during the intake stroke. That matters a lot at high RPM, where you’ve got less time available for intake. Optimized valve opening/closing times and intake runner lengths can set up, and take advantage of, pressure waves in the intake tract that will cram a last bit of air into the cylinder before the valves close, giving the best possible volumetric efficiency at the desired RPM.
Better in-cylinder flow structure. If you want to extract the most mechanical work from your burned fuel, you want most of that fuel to be burned quite close to top dead center, i.e. the start of the power stroke. So not only do you want to ignite it quite early, you’d like the whole burn to be quite short in duration. With two intake valves and idealized intake port geometry, you set up a “tumbling” flow in the combustion chamber, storing some kinetic energy in that rotating vortex. As the piston comes up to TDC and distorts this tumbling structure into something flatter, it also squeezes most of the volume into the pentroof-center of the chamber as the piston crown comes up very close to the head around its perimeter. Right around the time of ignition this organized tumbling structure breaks down into fine-scale turbulence that stretches and distorts the flame front, increasing its area and resulting in faster combustion. So for the same amount of fuel-air mixture coming into the engine, my current car makes more torque than my old car, i.e. it’s more efficient because of its faster burn.
Higher compression ratio. This facilitates extraction of a greater percentage of mechanical work from the combustion gases during the power stroke. My Q50 requires premium fuel, whereas my Cutlass didn’t. Higher octane is one thing that enables high compression ratio, but it’s not the only thing. Autoignition/knock/detonation has a time factor involved. If you compress a mixture to high enough pressure/temperature, high-octane fuel will detonate just as surely as low-octane fuel, but it’ll take a smidge longer to happen. The flip side of this is that if you can finish combustion more quickly - say, by running the engine faster or by using in-cylinder flow to your advantage - you can operate with a higher compression ratio independent of any increase in compression ratio that high-octane fuel might allow. Modern engine control tech also means that engines can be tuned to run on the ragged edge of knock for absolute best efficiency, because they can sense incipient knock that happens here and there and back off on the spark timing just a bit. That wasn’t possible 40 years ago, so engines had to be tuned to stay well away from the edge of knock throughout their operating range.
So that’s the big pieces of how you get 44% more torque per unit volume of displacement. The other factor in producing more power is bumping up the RPM at which the engine can be run. Better materials science, lubricants, and mechanical design have made this possible while still allowing the engine to last for a couple hundred thousand miles.
Ha! That was my second car. I always thought of it as a mini-caddy, with those soft velour seats!
Indeed. In the mid-late 70s what you had were essentially still engines from the 1960s, specifically those that could be made/tuned to run on unleaded, and to which early-tech emissions controls were bolted on. Not fully optimized, but a madate is a mandate. It took a bit for the US makers to catch up with the engine technologies mentioned in the thread (something that keeps happening in many sectors in the US, because of sheer size of market there’s often so much already-sunk financial, structural --and even emotional!-- investment in the previous technology that getting the new one deployed is a schlep).
Achates Power is working on this for vehicles. I don’t know how much detail they have in public-facing documents.
There doesn’t seem to have been any business activity since 2015, and the founder (Jim Lemke), while a legitimate scientist and inventor, died in 2019. I wouldn’t be holding out much hope that Achates Power is going to produce the next revolution in combustion engine technology any time soon.
Stranger
It was a great highway car, after you actually got on the highway. With a 0-60 time of 14+ seconds, actually getting onto the highway was kind of a white-knuckle experience.
I’ll meet with someone who used to work there later this month and will ask if it’s still active.
It was my dad’s lightly used company car, which I bought off of him in 85 or so. Was quite a contrast to the beaters and little shitboxes most of my fellow students were driving. And I’m not sure I noticed the performance a heck of a lot worse than the 72 Datsun B510 it replaced! For one thing, you couldn’t see the road through the floorboards once you pulled up the floor mats! 
Where do you throw your discarded RC Cola cans?
Stranger
Higher RPMs in a piston engine produce more HP. It means more combustion cycles per minute for the engine. If all else was equal the engine in your first car would have produced around twice the HP if it was running at 7000RPM also. All things aren’t equal, but many improvements allowed engines to run faster, on top of producing more power per combustion cycle.
I remember the first time I drove during a rain storm. Had no idea what was going on under my feet, until I realized water was sloshing in under the mats as I drove through puddles. Bought the car for $700 and IIRC later sold it to a friend for $300. When it crapped out on him, I understand he just left it on the side of the road somewhere.
But I digress/hijack…