Update 64 Faster Simulation, Cleaner Workflow, Tougher (Fairer) Challenges!

Unreasonably large open designs can run * moderately * now! Greetings, Hard Chip players!

Oh boy, it has already been 2 months since the last Dev Update!? I'm late😬! Here we go for a larger-than-usual update to catch up a bit. Simulation speed is always in search for more optimizations. Open Community Gauntlets is coming online alongside the continuation of Foundry Gauntlets. You can import chips directly into other chips from the Blueprint Library, the pause menu got a revision, the validation stack keeps getting stricter (and harder to cheese), and so many other things.

Let’s jump into what has happened from 0.1.7.128 to 0.1.7.147:

1. It’s Over 9000! Uncapped Tick Speed, Stabilization & Performance

Simulation now has new speed presets (1kT, 2kT, 4kT ticks/s) so you can push performance as you see fit, plus the ability to go beyond with an uncapped tick rate if your machine can handle it!

Want to see all the MOS and signal propagation sweetness? 250T/s is a good spot. Just want to see the result of the last optimization you made? Uncap those ticks/s and you’ll hit the Performance Report almost immediately!

Running your AND stage at 49103 ticks/s is of the utmost importance!

On the engine side, a lot of work went into doing less work: MOS states are evaluated only when a state change actually “touches” a MOS gate, large segment selection was simplified (and is much faster), and other optimizations were made. You’ll also see a live ticks/s counter (in extended info) and improved escape/input-lock behavior around the Performance Report windows. A few nasty stability issues got attention too (white screen of death, camera bounds, crash fixes).

Now usable to build colored segments mode!

Performance still has to go faster, though; simulation is plenty for the current 65 campaign and bonus stages, but you can already feel the limits on large stages like the 64-bit adder or with giant sandbox designs. The work continues as always, but now you can definitely reach new heights with the current state of HC!

⚡ Changelog:

• NEW It's over 9000 ticks/s!!! Uncapped tick speed + 1kT, 2kT and 4kT/s speeds

• PERF Simulation evaluates MOS state only MOS gate is being "touched" by a state change.

• PERF Large segment selection simplified and (way) faster run speed

• PERF Highlight points are shown progressively when more than 100, improved lifecycle for speed (tested with unreasonable 10 of thousands);

• UX Current live ticks/s show when extended live info option is on

• UX Escape behavior locks inputs for the editor temporarily, allow escaping/space/enter from perf report windows

• UX Color segment feature improved (and/or fixed? cause now more usable to actually build with).

• FIX White screen of death, evaluating/cancelling transformations separately;

• FIX Camera out of bound in overworld;

• FIX Crash to desktop tentative fix

• FIX editor "frozen" soft lock tentative fix

• FIX Rebindable Left Ctrl for upward build now working, Left Ctrl is not a reserved key anymore

• FIX Crash in benchmarks;

• MI Simplified loading schematics;

• MI Debug shaders fix

• MI Managing runner thread cancelling itself

2. Blueprint Workflow, Import Chips Into Chips + Better Probes

Blueprints got a new intended workflow: every time you create a blueprint, existing probes, inputs, and outputs are saved as probes into the new blueprint, always. Then, when you import, all probes are recreated by default. Effectively serving as labels on what is what, so you won’t have to guess if that wire is input A or B anymore! And when you’re done with those indications, you can remove probes