I asked Chat GPT to estimate power output on my engine build:


I'd like you to give me horsepower and torque estimates for an Oldsmobile engine I'm building. I can provide a lot of information.

ChatGPT said:

Absolutely — I'd be glad to help you estimate horsepower and torque for your Oldsmobile engine build. The more specific details you provide, the better the estimates will be.

Please include as many of the following as you can:

Engine Basics

• Displacement (cubic inches or liters)
  
• Block type (e.g. 455 Rocket, 350 Rocket, aftermarket)
  
• Compression ratio
  
• Cylinder heads (casting number, iron or aluminum, ported?, valve size?)
  
• Camshaft (brand, grind number, intake/exhaust duration @ .050", lift, LSA, centerline, hydraulic/solid/roller)
  
• Intake manifold (single or dual plane, brand/model)
  
• Carburetor or EFI (CFM rating or EFI details)
  
• Exhaust system (headers, size, mufflers, etc.)
  

Supporting Info

• Ignition system (stock HEI, MSD, timing curve)
  
• Fuel type (pump gas, race fuel, E85, etc.)
  
• Intended use (street, strip, circle track, etc.)
  
• Desired RPM range
  

Once you send that over, I’ll give you a realistic estimate of both peak horsepower and torque, as well as where in the RPM range they'll occur. If applicable, I’ll also suggest changes that could optimize your build.

Ready when you are.

You said:

427 cubic inch displacement, 4" stroke, 4.125" bore, stock block, 10.0:1 compression, CNC ported Speedmaster cylinder heads, 2.072" intake valve, 1.60" exhaust valve, Edelbrock performer RPM intake, Holley Sniper 2 fuel injection, full length headers, 93 octane fuel.

ChatGPT said:

Great start — that's already a solid list of specs, and it sounds like a very healthy Olds build with good street manners and strong midrange potential. Once you give the camshaft specs, I’ll be able to offer a much more complete estimate.

For the cam, please include:

• Duration @ .050" (intake and exhaust)
  
• Lift (intake and exhaust)
  
• Lobe separation angle (LSA)
  
• Intake centerline (if known)
  
• Type (hydraulic flat tappet, hydraulic roller, solid roller, etc.)
  

Let me know the cam specs and I'll run the calculations and give you estimated:

• Peak horsepower
  
• Peak torque
  
• Approximate RPM ranges
  
• Any bottlenecks or areas for improvement if desired
  

Ready when you are.

You said:

Hydraulic roller cam specs: Advertised Intake Duration 288, Advertised Exhaust Duration 294, Intake Duration @ .050" Lift 236, Exhaust Duration @ .050" Lift 242, Intake Valve Lift .555, Exhaust Valve Lift .576, 112 LSA

ChatGPT said:

Thanks! With those specs, we can now put together a solid estimate of your engine’s performance.

🔧 Engine Summary

• Displacement: 427 CID (4.125" bore x 4.000" stroke)
  
• Block: Stock Olds block
  
• Compression Ratio: 10.0:1 (pump gas-friendly)
  
• Heads: CNC-ported Speedmaster aluminum heads
  • 2.072" intake / 1.60" exhaust valves
    
• Camshaft: Hydraulic roller
  • Advertised Duration: 288°/294°
    
  • Duration @ .050": 236°/242°
    
  • Lift: .555"/.576"
    
  • LSA: 112°
    
• Induction: Edelbrock Performer RPM intake + Holley Sniper 2 EFI
  
• Exhaust: Full-length headers
  
• Fuel: 93 octane
  
• Target Use: Likely street/strip
  
• Target RPM Range: Midrange to upper RPM (2,500–6,200 rpm expected)
  

🧠 Performance Estimates (Realistic Ballpark)

Metric Estimate Peak Horsepower 510–530 hp @ ~5,800–6,200 rpm Peak Torque 530–550 lb-ft @ ~4,200–4,600 rpm

🔍 Analysis

🔥 Strengths

• Solid midrange cam: 236/242 @ .050" with a 112 LSA is a great compromise between street manners and upper-RPM power.
  
• EFI will offer better throttle response, drivability, and potentially tighter AFR control than a carb, helping torque and economy.
  
• Full-length headers and ported heads will let that cam breathe well and make full use of the displacement.
  

🧱 Potential Bottlenecks

• The Performer RPM intake is a good dual-plane but might start limiting peak HP beyond ~6,200 rpm. For more top end, a Victor-style single plane would help, though you’d lose some low-end torque.
  
• Stock block: You’re near the upper safe range of HP for a stock Olds block. If you’re pushing past 550+ lb-ft or over 6,500 rpm often, a girdle or aftermarket block is advisable long term.
  
• Head flow? If you have flow numbers from the Speedmasters, I can refine the estimate, but CNC porting + that cam = solid airflow in the 250–280 cfm range, which supports this power level.
  

⚙️ Suggestions (Optional)

• Consider a 1" open spacer under the EFI if clearance allows — it can help upper-RPM breathing.
  
• If you ever want to push into 570–600 hp territory, you’d need:
  • More cam (250°+ @ .050")
    
  • A single plane intake
    
  • Possibly more compression (11:1)
    
  • Higher flowing heads (~300+ cfm)
    

 

Then I asked:



With that engine in a 3,800 car with a manual transmission, 3.73 rear gears, street tires, and a driver with limited drag racing experience, how quick do you think this car would on a 1/4 mile drag strip?

ChatGPT said:

Great question — this is where all the bench racing comes to life.

Let’s break it down and estimate your ¼-mile ET (elapsed time) and trap speed, factoring in your real-world variables:

🏁 Car & Driver Summary

• Weight: 3,800 lbs (with driver)
  
• Engine: ~520 hp / 540 lb-ft torque (street-tuned)
  
• Transmission: Manual (assume 4-speed or 5-speed)
  
• Rear Gears: 3.73:1
  
• Tires: Street radials (limited traction)
  
• Driver: Inexperienced at launching
  
• Fuel: 93 octane pump gas
  

📈 Estimated 1/4 Mile Performance

Metric Estimate ET (Elapsed Time) 12.4 – 12.8 seconds Trap Speed 110 – 114 mph

⚠️ Why It’s Not Faster (Yet)

You're absolutely making enough power for 11.80s at 115+ mph in ideal conditions. But:

• Traction limitations: Street tires + manual + high torque = wheel spin city.
  
• Launch technique: An inexperienced driver will likely bog or spin the tires off the line.
  
• No trans brake or torque multiplication like in an auto + stall setup.
  
• Weight: At 3,800 lbs, the car is no lightweight.
  

🧪 The Math Behind the Estimate

Using the basic formula:
ET ≈ (Weight ÷ HP)^0.333 × 5.825
For your setup:
(3800 ÷ 520)^0.333 × 5.825 ≈ 12.15 sec (theoretical best with good traction)
But we adjust upward due to:

• Manual trans inefficiency
  
• Street tires
  
• Inconsistent launches
  

So real-world, 12.4–12.8 is realistic with some practice and street rubber.

🏆 How to Drop into the 11s

If you're chasing that 11.90–12.00 second zone, here's what would help:
Upgrade Impact Sticky tires (drag radials or slicks) ~0.5 sec drop — biggest gain Practice launches ~0.2–0.4 sec — especially with a manual Weight reduction Every 100 lbs = ~0.1 sec Line lock + clutch management Keeps burnouts and launches more consistent

Final Verdict

With your current setup:
✅ Expect ~12.5 seconds @ 112 mph, possibly faster with good shifting.
With drag radials and practice:
🔥 Low 12s or high 11s are definitely in reach.
Let me know if you want help with:

• Launching tips for a manual car
  
• Gearing or shift point optimization
  
• A power-to-weight calculator spreadsheet
  

This is already a quick street car — and it has even more in it.

 

Cam specs aren’t what I would do for this build. First off, why not more lift?

 

I accidently typed in the wrong exhaust valve size, here is the update.
Correction on the engine specs, it actually has a 1.68" exhaust valve.

ChatGPT said:

Thanks for the correction! Going from a 1.60" to a 1.68" exhaust valve is a meaningful change — especially for a big-inch Olds like your 427. Let’s revise the performance outlook with that in mind.

🔁 Updated Engine Component:

• Exhaust Valve: 1.68" (instead of 1.60")
  

🧠 What This Change Means

Increasing exhaust valve diameter improves exhaust flow and scavenging, especially on long-stroke engines with lots of displacement. This can:

✅ Benefits:

• Improve high-RPM cylinder evacuation
  
• Slightly increase peak horsepower
  
• Help with torque curve flattening past peak
  
• Improve Volumetric Efficiency (VE) overall, especially if the heads flow well
  

It’s especially effective with a healthy cam like yours and good flowing headers, which you have.

⚙️ Impact on Power Estimates

With 1.68" exhaust valves and CNC porting, let’s adjust:
Metric Previous Estimate Updated Estimate Peak Horsepower 510–530 hp @ 5,800–6,200 rpm 520–540 hp @ 5,900–6,300 rpm Peak Torque 530–550 lb-ft @ 4,200–4,600 rpm 540–560 lb-ft @ 4,300–4,700 rpmYou could easily gain 10–15 horsepower and 10 lb-ft of torque over the previous setup, all else being equal. The improvement is subtle, but real.

 

My current roller is on a 112 and I'm very happy with it's manners. I considered a 110 but I want to err on the side of driveability over power.

Lift, I can't answer. It's what Peyton suggested.

 

With all due respect to Peyton, why do a CNC’d head that flows more and more to .750, and only have mid .500’s lift?
I'm not saying lift it to .750, but there are things you can change to make the same power and still have good street manners.

 

The head/spring package is good for up to .600 lift, I don't know why the suggested cam isn't maxed out with an additional ≈ .040 intake and .020 exhaust lift.

 

AI sucks

 

Not really. Now you can sit at home, drink beer and BS ChatGPT and ChatGTP can BS you back. You don't need to get your buddies together for pizza, beer,and bench racing (BS). You can tell ChatGTP all kinda crazy chit and it believes you.

 

@ Mr Nick, Cant wait for you to get it on the dyno, really curious how good AI is at this.

 

Me too, hopefully all of the parts arrive by end of the year (less pistons), I can drop everything off to the builder and hope to get it back Spring/Summer 2026.

 

I'd be interested to see a 3 way comparison- track time/calculated hp, dyno readings, vs computer mumbo jumbo

So, basically, you just added your engine specs to the database for info to be pulled at the next request. You're seeing an average of every post ever made on the internet regarding an Oldsmobile engine with your specifications. I suppose it could be entertaining, but not really my bag. Build it to the best of your ability, then send it down a measured distance. You can't argue physics.

 

I don't think I added much info to the database with my request, since I didn't provide any feedback. Had I posted the results of an actual dyno pull along with my build combo, then I suppose ChatGPT could have "learned" from that.

The details of the build were spec'd by the Olds vendors I purchased the parts from. Then, I inputted the combo into ChatGPT out of curiosity. Yes the answer that ChatGPT provided is based off of searching the internet for a ton of data and then turning into a nice packaged reply. It may have even used some of your past posts as part of the data gathering process. The reply does seem pretty reasonable based off of what we know this builds end up making.

I'm very curious to see what my very limited experience drag racing a manual transmission does for the 60' and 330' foot times. I'll happily post my results when that day comes.

Here you can how consistently un-good I am, these are from 6-7 years ago before Rt 66 in Joliet pretty much shut down during the pandemic. This combo made 360/365 on a chassis dyno.