programming

[silent cloak]

they use less power

because ANNs are highly inefficient and power hungry

hence needing entire islands just to make them answer wrong on a test

[umbral thorn]

[sage crag]

beep

[silent cloak]

they dont have a concept of time either while SNNs do

[true hemlock]

data hungry

[silent cloak]

because they rely on "spikes" which are closer to human neurons

[silent cloak]

all for this

[uneven pulsar]

I need some help
Formate
([#fill here anything ])

[true hemlock]

real time learning not even possible on ANN since its heavily based on statistics

[silent cloak]

i hate distilled models so much man

[obsidian mantle]

what do these things do

run ais?

[true hemlock]

ANN fake neuron aswell tbh

[uneven pulsar]

Ann ???

[silent cloak]

dont get me wrong they are good for many uses, but for anything that uses NLG, hell no

[sage crag]

whole point of NN to begin with

different level of accuracy though

[silent cloak]

and they are incapable of actually thinking too unlike SNNs as they are continuous

[sage crag]

SNN "continuous"

[obsidian mantle]

is it just analog processor or what

[true hemlock]

continuous isn't the only determinant for capability of thinking though

[silent cloak]

of course not but they are capable of actually learning instead of feeding itself with it's own context over and over

[uneven pulsar]

Subprocess.run([])

[sage crag]

it interesting because it obviously discretised, i think human brain discretised too ne

[silent cloak]

"thinking" in traditional ANNs are just chained prompting

which is cringe

[true hemlock]

if i may ask how do you roughly implement synaptic plasticity in your case

might be interested with sharing my research

[sage crag]

ANN just artificial neural network kinda generic term ne?

[fast pagoda]

it just learns how to prompt itself

"learns"

[obsidian mantle]

how do you even discover this

first time i hear about it at all

[sage crag]

deep learning still learning ye

minor semantic difference

[obsidian mantle]

do you need some special motherboard for this

or is it connected via usb

[sage crag]

talking about neural network architecture

if you mean neuromorphic hardware, its specialised

[obsidian mantle]

yes the hardware

[fast pagoda]

also ive returned from JEP106

mfw i dig throuhg the module manufacturer thinking the thing actually gave me some info and the "module manufacturer" is listed as bank 5 0xCD which is

drumroll please

g.skill

[silent cloak]

there is no "learning" the model weights are left untouched.

[fast pagoda]

and i know those mfs dont make modules

[silent cloak]

modulated STDP

kind of a newer one to the table

but it shows promise

[fast pagoda]

i was talking about training phase when "learning" to use "thinking"

[sage crag]

probably means learning as training

[true hemlock]

fuck sleep, im more curious with how others that are also involved in this field actually also implement this

[silent cloak]

thats the difference

[fast pagoda]

konii got me

[silent cloak]

SNNs are capable of learning post training

[sage crag]

ye ye me know

synaptic plasticity

[fast pagoda]

yeah that is super coolio i was just aka repeating the same thing you said about traditional models and their "thinking"

[obsidian mantle]

??? are you running it in simulation in your normal cpu or what

do you have actual processor of that kind

[silent cloak]

u can run SNNs on normal hardware its just not as efficient

[obsidian mantle]

what is the "research"

[fast pagoda]

the hardware itself does not need to be bespoke

it can be

[obsidian mantle]

[silent cloak]

the research is essentially math which isnt the most fun

I have dopamine as a factor in my model because i based it on DA-STDP

[sage crag]

lacking the entry level ML knowledge to understand me think

[silent cloak]

which is a type of modulated STDP

[sage crag]

need go and watch some videos

[fast pagoda]

that's how GPUs are where they are, there are ways to make specific accelerators for these jobs that beat GPUs which are of course general purpose, but they brute force it and have software support so running them is what we do even tho the hardware isnt really built for it (well, increasingly it is now but)

[silent cloak]

they arent a GPU technically

[fast pagoda]

yeah im talkin like running any model in general and just general purpose hardware vs not

[silent cloak]

NVIDIA malds at the potential of SNNs

[fast pagoda]

kek

[obsidian mantle]

konii your entry level is like top 0.00001% level i aint watching 50000h of videos

nvm then i'll just ignore this topic

[sage crag]

ne not ML knowledge

[silent cloak]

@true hemlock https://arxiv.org/abs/2504.05341

[obsidian mantle]

there are too many things in this world i havent heard about

[silent cloak]

a good read though not the only paper on that version

[sage crag]

my ML knowledge is pretty introductory

me systems programmer not ML

[fast pagoda]

just go watch 3blue1brown's intro on neural networks if you want a basic idea of how at least pre-transformer style models function at a conceptual level as well as how transformers themselves work (again, conceptual and visual) in comparison

not much material out there about SNNs designed for someone with no prior background in ml or at least a passing interest to have the baseline of understanding

[obsidian mantle]

i think i watched it

[fast pagoda]

but did you WATCHED it

[sage crag]

with SNNs, is it possible to represent less important parts of the network with fewer states? i.e. reduced precision

me not know

[obsidian mantle]

maybe i watched something else

idk

i cant connect that to what you are discussing here

[silent cloak]

i mean weight quantization is always an option though there are other methods too

[obsidian mantle]

since when are they making fucking processors that are based on that and which do work why has nobody ever mentioned it anywhere

do they suck so bad its just like a toy project at its peak

[silent cloak]

though u also have the option of temporal based thanks to SNNs being temporal based

[sage crag]

isnt the typical issue with quantisation is its global and quite lossy? additionally its a post-processing step which isnt really suitable for learning from

which, im not sure how that affects snns but ye

[silent cloak]

i mean its pretty much only used to shrink ANNs because of how hungry they are

[sage crag]

yee

[silent cloak]

SNNs are already far lower in power requirements but u can still employ a number of methods

[obsidian mantle]

what does it get as an input

if not numbers

[fast pagoda]

@silent cloak what you think about this https://www.mdpi.com/2673-4117/6/11/304 because it got mailed to me (by fuckin grok lmao who i have send me ml news every day) this morning literally

[sage crag]

i was under the impression SNNs dont take input in the same way traditional neural networks do

[obsidian mantle]

all neurons get input simulatniously?

[fast pagoda]

they get input ...over time

[sage crag]

@true hemlock

[silent cloak]

ANNs dont have a concept of time

SNNs do

[sage crag]

[obsidian mantle]

so it depends on time too

[silent cloak]

and then each neuron's signal decays over said time

so its similar to how the human brain works

[obsidian mantle]

thats it?

seems useless

[fast pagoda]

you could say that's it if you simplify how our brains work too

[silent cloak]

no because its not firing everything at once

like an ANN does

[fast pagoda]

but then you're not talking about the wobbly signalz and shit

[silent cloak]

so its far less power hungry

[true hemlock]

yeah they generally don't. depending on implementation, in my case its stimulation based so depending on which cortex net loop picks it up then dynamically propagates it to other loops

[silent cloak]

and because the signal propagate u have many options of learning beyond the initial training if u employ neuro plasticity

normal ANNs cant learn by doing, only by training

[sage crag]

kinda like sensory neuron in brain?

[silent cloak]

sure u can fill the context with history but that just increases the demands more and more

[true hemlock]

the first layer yeah, then second layer and so on but all the "neuron" per layer received it at once. its literally just linear algebra though

[sage crag]

quite literally, adding, multiplication, and the application of a single function per weight

[obsidian mantle]

this is regarding ann? and for snn?

[true hemlock]

not for snn, at least on some implementation

the basic implementation of snn is still linear

just works with timing based firing with thresholds

[silent cloak]

btw for an example 2-bit operations on neuromorphic hardware is generally 16x faster than 32-bit

4-bit 8x etc

they dont demand as much

[obsidian mantle]

and this field sucks so bad nobody uses it for anything real yet

or do they?

[silent cloak]

some do

its experimental

just like ANNs were

i only found vedal originally because AI as a subject was obscure back in the day

though machine learning still had its uses

[fast pagoda]

they are new

ish

[silent cloak]

the first Ai related thing I made was LVQ network

[fast pagoda]

at least in the way that people are excited about and researching and implementing them

[silent cloak]

then i moved to NLU and to NLG

[silent cloak]

big NVIDIA researched them years ago but they cant make as much money selling GPUs if it pops off

u can find some old presentations and papers they did

[fast pagoda]

tbf they probably kinda sucked compared to sota of like, unet and other CNNs etc at the time which had more interest and support around them even then

snns as a concept have been around a hot minute though yeah

i dont know when this was so idk what they would've been comparing them to

i am gonna go try to find paper though kek

[silent cloak]

one of their presentations they did i think was in 2019?

[fast pagoda]

oh so that's against transformers and like unet type shit at that point

[true hemlock]

its not efficient back then. we're more forward with SIMD based compute so CNN and later on some other popular ANN pops up faster

also people gets the comparison wrong often

[silent cloak]

thankfully we dont need either for these

[fast pagoda]

makes sense

[true hemlock]

your typical SNN would suck at most data generation workload

[silent cloak]

they use spiking neural circuits in some of the hardware to do two's complement and binary matrix multiplication

[sage crag]

is a rig with my specs decent enough to get a toy SNN running?

[silent cloak]

and they honestly are able to work just off of INT2 for a lot of processing

[sage crag]

i have no gpu, 32 gigs ddr5, and a ryzen 5 7600

[true hemlock]

they're much better off doing decision making

if im ngl

llm suck at decision making

[silent cloak]

indeed

[sage crag]

you can offload the decision to daughter networks though

[true hemlock]

considering that it has no concept of timing

[silent cloak]

or adaption

[obsidian mantle]

what does concept of timing have to do with all of this

[silent cloak]

SNNs are very popular atm for visual and audio processing though

[fast pagoda]

temporal information in models and using it for anything is so hard for me to understand somehow

[silent cloak]

SNNs are event driven

[nocturne olive]

Neuro when given the option to do nothing or enter a menu when she's about to win a game: enters menu for 2 minutes after the game has ran for 10 seconds

[silent cloak]

and timing is needed because the spike signal must decrease over time

[sage crag]

"annoying bubbles"

[silent cloak]

if i had a drawing board i could explain it better

[sage crag]

evolve devolve evolve devolve

[obsidian mantle]

does each neuron have its own "decrease rate" or they do it all in the same rate

[sage crag]

SNNs wouldnt struggle so much with that loop she got into in plague inc

because they retain context

[hearty notch]

excalidraw dot com

[silent cloak]

they have their own decay

[hearty notch]

draw it share it

[silent cloak]

thats their "weight"

[fast pagoda]

SNNs are kinda like async while traditional networks are more continuous and linear

[silent cloak]

there are a few different neuron designs rn though

[fast pagoda]

am i horribly wrong here

[silent cloak]

because its all so experimental

[true hemlock]

on one side with ANN you practically have a complex network of matrix multiplication to determine the data to be predicted. yeah these models don't have the capacity to think nor have the concept of time, its just predicting data as quick as the computer can do

on the other hand, SNN are more related to biological neurons. the firing are more timed

[sage crag]

well, you can think of it more like, a traditional network is inferred all at once

whereas an SNN is more "partial" or "async"-like you said

[true hemlock]

there are cases where SNN model creates feedback loops and it sometimes works as a timing sensing mechanism

[obsidian mantle]

so it depends not only on values of input, but also on the time and order these values appeared?

[fast pagoda]

right so it's kinda idle until it reaches activation and as a function of that event firing sort of approach, it inherently includes time in its computation of a result

[sage crag]

brain IRL do this too

[sage crag]

bees have sub-second clock in their head

why do

[hearty notch]

have you ever seen high speed cam of bee wings they propel upwards with both flaps it looks so cool

[true hemlock]

this is the SNN implementation as a feed forward for a CNN isn't it

[silent cloak]

showing some of the comparisons

[obsidian mantle]

what is that "IF" thing

[silent cloak]

whats ur favorite neuron as of lately

Leaky is a popular one

[nocturne olive]

Scam bot, call mods

[sage crag]

yee

look at this

[true hemlock]

fair, SNN are pretty competitive in computer vision somewhat and ANN also works so its a proper comparison for a specific usecase

[obsidian mantle]

and who processes that "bunch of spikes output" into actual data

or its only midlayer drawing

[fast pagoda]

encoder/decoder?

is that still a thing for snn

[silent cloak]

yeah most papers are related to audio and visual

which is kind of how ANNs started out in research

[true hemlock]

pretty fun.

[fast pagoda]

dont talk about that, quack will be angy

it's over

[sage crag]

🦆

[hearty notch]

it's so over

[sage crag]

wasnt this their first big popularity thing

in a computer vision contest

[fast pagoda]

i guess it makes sense to pursue something initially that is exceptionally obvious to the layman "hey that's getting better!" and is also easier for us to intuit ourselves

[sage crag]

or benchmark

[silent cloak]

binary neuron

just accumulate until threshold pretty much

[sage crag]

neuron doesnt output until it goes over threshold

[obsidian mantle]

when does it stop the output? it has decay too?

[sage crag]

and its activation continuously decays until it hit threshold

[obsidian mantle]

once threshold is reached

[fast pagoda]

similar to an action potential

[sage crag]

typically output is an instantaneous event

it doesnt "last" for any time

[obsidian mantle]

ohhh so it holds the output while it has enough intput

[fast pagoda]

that's how our neurons work w/ action potentials and the threshold they have to actually fire

at least in part

[obsidian mantle]

how the hell can this be applicable to anything

[sage crag]

close enough

[silent cloak]

alot of different neuron types are being experimental with rn in these models

[obsidian mantle]

you need some weird input converter and weird output decoder idk

i have jpg image what do i give it

[fast pagoda]

regular models have that too

[sage crag]

yee

[fast pagoda]

encoder/decoder

[sage crag]

you think regular vision model can accept 1920x1080 image?

lol

[fast pagoda]

although i saw a paper where they were getting damn close to a practical way to not have the enc/dec step

[sage crag]

terrible use of input neuron

need preprocessing

[obsidian mantle]

oh so it looks like low resolution black and white

[sage crag]

depends on the preprocessing

[fast pagoda]

lot of the challenge in any ml has been getting the information the model needs to operate on

to the model

in a way that is helpful

[silent cloak]

ok let me explain how the basic neurons with these work simply

[sage crag]

more typically, an SNN could examine parts of an image at a time without losing context

[silent cloak]

imagine a bucket

fill it up from the other neuron signals moving in

[true hemlock]

close, but this only explain the basis of a neuron (with also missing some other details, though. such as specific axons and concentration on a group of stimulated dendrites)

[sage crag]

whereas a normal ANN couldnt

[silent cloak]

the water will slowly leak as the neurons forget

but if the water reaches the top, it shall fire

[fast pagoda]

yeah im abstracting it a lot to basically be this

[silent cloak]

continuing the cycle

[obsidian mantle]

and that "can learn on the fly" part? how does that work

[silent cloak]

neural plasticity

different methods exist

STDP being the most famous

[true hemlock]

this is where people getting confused with
you don't input raw data, these models are meant to work with stimulus based (separate encoding also exist)

[sage crag]

dont they typically have rapid decay after firing too?

[silent cloak]

Depends on the neuron being used

[sage crag]

[silent cloak]

again there are a few different types rn

[sage crag]

yee makes sense

[silent cloak]

there is no standard yet

leaky is a popular one

[fast pagoda]

they adjust the activation threshold for next time via some magic that accounts for how long the activation took

[obsidian mantle]

so it looks at the output and can change itself depending on the output... what stops us from doing the same in ann?

[true hemlock]

my research institution even made some new types of neuron

[fast pagoda]

catastrophic model collapse

[true hemlock]

classified, though

[silent cloak]

its not temporal based

[fast pagoda]

is what happens with a regular modelif you adjust the weights

on the fly

too much

[true hemlock]

:TOSbutNDAemote:

[fast pagoda]

same

[obsidian mantle]

does SNN revert to original state overtime?

[silent cloak]

essentiall with STDP connections between neurons get stronger when the first neuron helps cause the second to fire

and it gets weaker when it doesnt

[true hemlock]

if you do that shit on an ANN it'll overfit and learn noise instead.

[silent cloak]

yeah and they use VALUES not spikes

which is also why they are inefficient

lets say ur on a phone call

the ANN will continue to process all the empty information

[obsidian mantle]

snn learns on the fly from the input?

[silent cloak]

they also use energy when nothing important is going through

SNNs are also generally binary

[fast pagoda]

yep thats why dense ass models are horrible to run

[sage crag]

ohhh quuck i wanted to show you a funny thing that i caused chatgpt to do

[silent cloak]

0 or 1 while an ANN will send 0.7, 0.8, 0.85 etc

[sage crag]

https://chatgpt.com/share/6904fce0-7e30-800c-97f4-9c1c0ef4a1d8

take a guess at how many semicolons are in that output

about 17kb of them apparently

[fast pagoda]

you still have to do the non-firing state updates in the neurons in a SNN running on something that is NOT neuromorphic right

so it's not quite as efficient

[true hemlock]

im not going to downplay ANN either and be the loud minorities of SNN folks though. gonna make it clear that SNN have different usecases.
you don't generate semantics token vector with these model to generate text, nor you able to diffuse noise from these model.
dunno how to make it simple, but i guess you can say SNN are far better with timed decision making, while ANN are more suited for generative usecases

[silent cloak]

the time steps but the neuron wont get anymore input

think of it like this

an ANN is contantly talking even when nothing of importance is happening

while an SNN will only talk when needed and put the phone down after

[fast pagoda]

and they only do a couple steps for SNNs too anyways even if you're calculating some updates per non-firing synapse

[hearty notch]

ITS NO NEURON NOVEMBER GUYS WTF

[silent cloak]

they are more simple mathmatically so they still dont draw nearly as much power

[fast pagoda]

it's no nVedal november actually

[sage crag]

i wonder if its possible to run an SNN on analogue hardware

[silent cloak]

they dont need large values

[fast pagoda]

brain is analogue hardware running an SNN :^)

[sage crag]

need plasticity so need specialised version ne

[silent cloak]

neuromorphic chips usually use a mix of digital and analog

[sage crag]

brain is not analogue

[true hemlock]

right

[sage crag]

make sense

[hearty notch]

why wouldn't you consider it analogue it has materials that aren't in standard digital

[sage crag]

biological not analogue

[fast pagoda]

that's too far into the weeds, actually i dont know what most people would even think of the brain as if they had to classiffy it

[obsidian mantle]

how is brain not analogue

[hearty notch]

what's your def of analog

because mine is just not digital

[sage crag]

analogue computation specifically

[fast pagoda]

analogue to me is just continuous rather than discrete mostly

[hearty notch]

I see

[sage crag]

brain is definitely not continuous

[fast pagoda]

you are correct there for sure

well

as a system it kinda ends up being

[silent cloak]

neurmorphic chips run async too which is nice

[obsidian mantle]

why not

[true hemlock]

speaking of which my research institution has quite done slightly outside the general meta for snn rn
i got a question, by any chance is there any publicly known spiking neuron that needs occasional synaptic regulations
couldn't find any of these yet in public papers but in our variation of snn we do have this shit

[silent cloak]

because they just run off of spikes and dont need a global clock

[fast pagoda]

but individual firings and such are what like 10 hz or something

[sage crag]

evolution ne

[obsidian mantle]

how is this connected

[sage crag]

dont ask me ask common ancestor of life with brain

[hearty notch]

ultimately nothing's continuous if you consider planck time the fundamental resolution of time

[sage crag]

planck time isnt resolution of time

[silent cloak]

also no complex floating point matrix multiplication

which is always great

[sage crag]

there is no evidence time is not continuous

[true hemlock]

this whole channel is lunatic

lmao

[silent cloak]

of course

[true hemlock]

one side we have 2 SNN bros

[sage crag]

planck time is just the hard limit for measurement of time

[true hemlock]

other side is bunch of people who never worked with ml fields

[silent cloak]

lmao

[sage crag]

yeee it me

[true hemlock]

some has, but on ann

[sage crag]

im the clueless

[true hemlock]

then you have random philosophical statements outta nowhere

[obsidian mantle]

how is brain not continuous

it is continuous

[sage crag]

ne, ye not know how brain work

[silent cloak]

without neurmorphic hardware SIMD does a good job though

[true hemlock]

this is so #programming piled

[fast pagoda]

neurons generate impulses but they are not continous

a neuron is putting out discrete events

[silent cloak]

yup

[hearty notch]

I mean even if you're considering every neural interaction discretized there's meaningful stuff happening between those discrete moments no

[silent cloak]

some neurons arent really firing at all either depending on the situation

while an ANN is firing EVERYTHING

[sage crag]

it amazes me how many people have this misconception btw

[fast pagoda]

that's what i was saying, the system as a whole kinda operates as if it were continuous and certainly our own perception is continous or feels that way so it makes it hard to intuit

[sage crag]

planck units arent anything to do with fundamental limits of physics

they are about useful numbers for us to use in computation

[true hemlock]

then you have the physics bros

[sage crag]

[hearty notch]

it's a fundamental limit of measurement which is a fundamental limit it's not about our instrumentation it's about what's literally possible

[true hemlock]

don't get me wrong though this chat is entertaining

[obsidian mantle]

what even is time

there is no time

its subjective

[silent cloak]

they are talking about a dimension as if the mathmatical representation is the part that matters

💀

[sage crag]

not relevant ne

[obsidian mantle]

you cant prove it if its not measurable

can you

whats the point

[hearty notch]

time is measurable idk where you got where it's not

[silent cloak]

i mean

technically its not

[obsidian mantle]

i mean below plank time

[silent cloak]

just a representation

[sage crag]

there is no absolute time either

[silent cloak]

u cant measure time from the third dimension

[sage crag]

the speed of time changes when you move

[hearty notch]

planck time is on the order of e-44 seconds lmao we can do plenty above that

[silent cloak]

because ur stuck in that axis

[obsidian mantle]

the sky is blue

[silent cloak]

its like a 2D creature trying to measure the third dimension

how can it?

[sage crag]

i dont want to explain relativistic effects today so pls go explode

[silent cloak]

it cant reach up now can it

it cant throw a ruler onto it

[true hemlock]

is space continuous

[obsidian mantle]

konii is there anything you dont know

[sage crag]

yee i dont know where i left my headphones

[obsidian mantle]

can you go solve world hunger or smth

[sage crag]

there is no evidence that space isnt continuous

[fast pagoda]

neurons if they were continuous would not have a refractory period, it would not have a break to reset the gradients of like sodium and shit that cause it to fire

[sage crag]

but the answer is "we dont know"

[fast pagoda]

it somehow took me 8000 years to make this a white background

[hearty notch]

neural firing is discretized is still not a claim the entire brains function is but yea it's important to understand the model ig

[silent cloak]

we are talking about SNNs and dimensional physics

very interesting

[fast pagoda]

average programming chat

[sage crag]

nee its too chaotic and uninformed i feel like im going to die

[silent cloak]

i love both so im chill

[sage crag]

must be how quuck feel when i ask question about ML

[obsidian mantle]

i kind of want to become a known guy here since it looks like this channel is bedlam of future world ruling CEOs

[fast pagoda]

i find nothing wrong with knowing yourself to be uninformed and thus speaking to people who do clearly have experience in it, to try to understand better

which is why i do it constantly

[sage crag]

people with knowledge dont become CEO they become hermit

[fast pagoda]

they become woz

[silent cloak]

u have to be old and stupid to be a CEO these days

[fast pagoda]

or young and evil that's a good one too

[silent cloak]

and always make sure to ignore ur customers

in favor of shareholders

[silent cloak]

if u fuck up, you must double down

[sage crag]

nvidia at $5t

[silent cloak]

and insult the customers

[olive sable]

Goodmorning

[silent cloak]

let them know that u dont care about them and hate them

[sage crag]

samuel vanuel manuel

[silent cloak]

then u will be the perfect 2025 CEO

[fast pagoda]

i hate to bringit up but that's a failing of the fundamental purpose of the company being serving the shareholder above all else and not doing so would be a very bad for the company who is controlled by those shareholders

not that i think it's a good thing

[olive sable]

koniiiiiiiiiiiiiiiiiiiiiiii

[fast pagoda]

it's just kinda they kinda gotta (should not work that way)

[silent cloak]

yeah but the shareholders invest based on the involvement of consumers

[sage crag]

publically traded company is run by shareholders

[silent cloak]

if there are no consumers there are no shareholders

[olive sable]

What

[fast pagoda]

well, shareholder want the funny number to go up NOW not in 15 years

[true hemlock]

how the fuck did we went from SNN, ANN vs SNN, analog, biology, quantum physics, philosophy and then to business and capitalism

[silent cloak]

being a shareholder is investing in a company in the hopes that it generates profit from its product/work

[sage crag]

late stage capitalism is about making number bigger ye

[sage crag]

adhd

[fast pagoda]

and so everyone gets sacrificed except ol benjamin franklin on his hunned

[fast pagoda]

unironically this

[sage crag]

[silent cloak]

we are all markov chain chat bots

[sage crag]

not me im a tv

[fast pagoda]

reminds me i have something to consume to make me a good boy for society that i've neglected

[olive sable]

Idk man i just got here

[sage crag]

samuel vanuel manuel

🔺

[olive sable]

I slept for 14 hour cuz im walking like gregory house with my injury

[silent cloak]

based reference

[sage crag]

how you even get injured you dont go outside

[silent cloak]

phantom injury

[olive sable]

I do go outside

Im a college student

[hearty notch]

that's good it probably contributed to healing faster

[sage crag]

i cant believe this

[hearty notch]

get well soon!!!

[true hemlock]

fun fact, axons stimulation only stimulate relatively close local groups of dendrites while being quite far from its own neuron cell
this makes cortex loops possible and sort of giving the branched propagation

[silent cloak]

https://www.youtube.com/watch?v=cXA0Sgqz304

i forgot about this video

[fast pagoda]

alright tringle man let's get your take on the implications of the planck limit on whether or not the brain is digital or analogue or some other 3rd thing (biological system which can fit both in some ways or otherwise)

[sage crag]

stinky tracking

[silent cloak]

accurate ANN representation

[rigid snow]

pp

[silent cloak]

my bad i have a thing that usually removes it

[sage crag]

yee must be new one

ive seen it show up

[fast pagoda]

i was about to mention having one of those

[sage crag]

not in any filters yet

[olive sable]

Idfk. I thought the electric oulses were "digital" and hormones and stuff were analogue

[hearty notch]

I looked up some stuff about planck time and the precise formulation is it's the unit of time below which we are unable to reconcile the math of GR and QM and therefore have no models to predict anything

[sage crag]

even hormone is discrete

[olive sable]

Im not 1 biologist

Idk

[fast pagoda]

lmao tthe hundreds of doctor agents

[sage crag]

biochemistry is interesting, the way hormones work typically is by binding to a molecule and altering it in some way

[rigid snow]

[hearty notch]

definition of analog vs digital is a fight I may be willing to relitigate but I need to look up history of the idiomatic use of analog

[olive sable]

I know more about dna then hormones, dna is just way cooler imo

[sage crag]

ne me disagree

dna simple part of biochemistry

[olive sable]

And i only learn stuff im interested is cuz other info doesnt stick

[sage crag]

biochemistry go much deeper

[fast pagoda]

dna and hormones are pretty intertwined

[silent cloak]

DNA fans when their telemeres decrease

[uneven pulsar]

I must be song writer

[fast pagoda]

inasmuch as your gene expressions directly influence your response to any hormone

[sage crag]

you have 22 mitoses left before you turn into dust

[uneven pulsar]

Idk about biology

[hearty notch]

epigenetic mechanisms in DNA are cool. neurotransmitters are cool because I'm a huge fan of acid and that's how we discovered the phenomenon of the serotonergic system. GO ACID.

[silent cloak]

i dont wanna be taller i want my telemeres to be taller

[true hemlock]

among all the diverse topics from different fields randomly spouting out, here we are with outside randoms randomly make noise

[olive sable]

New year countdown but for mitosis

[true hemlock]

truly #programming piled

[silent cloak]

could probably make a clock like that

[true hemlock]

this is just the peak of this channel

[silent cloak]

based on estimates of course

[fast pagoda]

Mediterranean diet

[uneven pulsar]

TELL ME U GUYS
HOW MUCH URE GPU NEED TO TRAIN ?

[silent cloak]

i have a variation of it and keto

[hearty notch]

long walks

nice

[hearty notch]

I'm shortening my telomeres like idgaf I don't wanna live to 70

[sage crag]

[olive sable]

1 epoch, idk

[silent cloak]

A normal human cell can undergo between 40 and 60 mitotic divisions before it stops dividing and undergoes apoptosis

so it would be easy to estimate how many before death

based on average cell lifespans, general mitosis, and cell count

[obsidian mantle]

why do you remember all this

[sage crag]

unless youre henrietta lacks

[olive sable]

Can i just get more cells from those stem fuckers in my spine?

[obsidian mantle]

does your brain not forget things that you dont think about/encounter anymore

[hearty notch]

I had to read that book for a health data elective, the immortal life of Henrietta lacks

[fast pagoda]

lol the hela cells are fucking creepy to me

[hearty notch]

it was a good elective

[fast pagoda]

any of the immortal cell lines

[sage crag]

i just dont remember anything outside of random information

[silent cloak]

thats called object permanence

[sage crag]

forgot to eat for 10 hours ne

[nocturne olive]

Depends on the model

[fast pagoda]

it's super cool and useful for research but damn

[uneven pulsar]

Google server gpu spec
4²

[olive sable]

Same lmao

[sage crag]

yeee

[maiden geyser]

№2

[olive sable]

I need to take my meds too

Brb

[fast pagoda]

to have your cells exist 70+ years after you've died

that is wild

[silent cloak]

imagine if telomeres didnt suck

[sage crag]

maybe in a century she will be revived

[sage crag]

some species can regenerate their telomeres

[silent cloak]

lucky

[fast pagoda]

a Helamunculus

[uneven pulsar]

ASK QUACK TO DO FINE TUNING

[fast pagoda]

im so fucking funny bro

[obsidian mantle]

this conversation feels like from another world where everyone knows everything for some reason

[sage crag]

yee enzyme called telomerase

uuh

lobsters have it

[silent cloak]

me programming SNNs with telomeres

[hearty notch]

afunyun which char is your pfp

[silent cloak]

so they eventually die like people

[sage crag]

planned obsolescence

[silent cloak]

neurons do have telomeres

in humans

[sage crag]

iSNN

[nocturne olive]

That i person is so weird

[sage crag]

ye

[fast pagoda]

mailman/Incidental 154/guard 1/policeman 2

[true hemlock]

in our (my research institution) case, the synaptic plasticity does multiple things actually. strengthen "axons" into local dendrite group, shift to a different group, split and refactor groups. the signal propagation isn't considered as synaptic plasticity though in our case
we modelled our SNN variation heavily based on advanced research in neuroscience aswell
could be useful if you're planning something similar but in most cases of SNN only research its not as applicable.
@silent cloak

[uneven pulsar]

My one guy has name call deleted user in other guild.id

[hearty notch]

ty

[true hemlock]

time to dip

[sage crag]

me donate my brain to quuck.ai after i die

dont send ninja after me

[silent cloak]

its heavily based on real neurons so yeah

i had to read up quite a lot

[nocturne olive]

As usual, a completely nonsense sentence

[hearty notch]

oh ok I knew it looked maybe spongebobby I thought Simpsons at first but this makes sense yea the oval eyes

[sage crag]

i think its just ESL

[silent cloak]

ANNs are more math than biology while SNNs are more biologically focused

[fast pagoda]

he tries lol i don't mind the entropy he injects

[uneven pulsar]

Snn hmmm

[olive sable]

hi

[sage crag]

also kinda similar text mannerisms to that 505 guy

you know the one

[uneven pulsar]

Hello

[olive sable]

i am on desktop now

[sage crag]

not seen him talk for a while

or was it 404

not sure

[silent cloak]

though ignoring all the other work that goes into it regarding the math of course

[true hemlock]

not sure if im willing to invite cipher somewhere for more of our snn research but

actually nah

night drunk rn

[sage crag]

take out to dinner first ne

[olive sable]

if we have antibiotics, then what are biotics?

[fast pagoda]

you

and me

[uneven pulsar]

Atleast I got idea tho

[sage crag]

garlic ne

[olive sable]

[olive sable]

mods

spammer

[hearty notch]

oops

it didn't go away

I thought it wasn't going through

[obsidian mantle]

[olive sable]

[hearty notch]

on mobile

[sage crag]

Dulbecco's Modified Eagle Medium

[obsidian mantle]

of course

[olive sable]

what

[fast pagoda]

average red goop

[obsidian mantle]

wait let me try something

[sage crag]

biotic just something that supports the growth of cells

DMEM is a biotic

[fast pagoda]

soup for shit to grow in a petri dish

[olive sable]

ew

[silent cloak]

programming chat is just spinning the wheel on topics

[sage crag]

weird example ne but funny

[hearty notch]

probiotics and prebiotics have been been fads for awhile

[obsidian mantle]

ahem.. so there is this problem where we cant travel faster than light, but i heard there is some trick how its actually possible, have you heard somehting about it @sage crag

[hearty notch]

I see them extra because california

[uneven pulsar]

Well any one has idea to mix multiple image to one ?

[olive sable]

we're all just nerds with similar interest i think

[fast pagoda]

if you get enough people chatting at once this chat cycles throuhg every possible STEM/Philosophy/brainrot/gibberish topic imaginable

[hearty notch]

I'm ADHD as hell and will talk about anything I know about

[sage crag]

there is no way of travelling faster than light, but there may be ways of altering space such that you travel less distance to get to a destination, hence creating the illusion of moving faster than light

no actual evidence of such a method existing though

[obsidian mantle]

oh yeah of course how do i do that

[fast pagoda]

same
and also things i barely know about

[hearty notch]

same

[sage crag]

invent warp drive that bends spacetime

dunno no one did it yet

[olive sable]

welp, im still not able to walk but it hurts less. i might skip school tomorrow

[obsidian mantle]

can you do it for fun

[fast pagoda]

ALCUBIERRE DRIVE

[uneven pulsar]

Use #general-network chat

[fast pagoda]

JUST 10 MORE YEARS GUYS

[sage crag]

ne too expensive

[obsidian mantle]

but you know how

[sage crag]

[flint dome]

until we find a way of bending space to our will we're stuck with the "light barrier" much like we had the "sound barrier" before yeager

[sage crag]

speed of light isnt barrier

speed of light is fundamental limit on how fast information can travel

it insurpassable

[fast pagoda]

the speed of light is the great filter 🤯

[sage crag]

quite possibly

[olive sable]

every now and then konii gets a new word in her vocabulary, and now it is is "ne"

[hearty notch]

not mutually exclusive

[sage crag]

wrrr

[fast pagoda]

ne has been increasing in frequency for a stinky minute

[sage crag]

https://cdn.discordapp.com/attachments/1126496895246807123/1433049661270261871/ShouldICelebrate.gif

[uneven pulsar]

Do u know we are offical stage 0
In technology

[olive sable]

wrrr

[hearty notch]

ne ne, kiiteru?

[sage crag]

no

[flint dome]

if i remember, dont you approach infinite mass when you get near the speed of light?

[olive sable]

yes

[uneven pulsar]

Stage 1 has able to use whole sun engery

[hearty notch]

yep

[fast pagoda]

given that kardashev is based on energy consumption

if we want to be as high as possible

give NVIDIA more money

[true hemlock]

diagnosed Autism + ADHD goes hard

[sage crag]

pretty much

[fast pagoda]

jensen will figure out harnessing the sun for a megaslopper 9000

[uneven pulsar]

Nivida need to make
More gpus 🙂

[sage crag]

there's more too it than that though

[true hemlock]

Schrodinger's state between hyperfixation and lack of particular focus

[flint dome]

thats why i was saying it is sort of like a barrier to getting fast enough

[sage crag]

otherwise you would still be able to get to the speed of light in a vacuum

[olive sable]

E = mc^2. so mass becomes big

[hearty notch]

barrier is a fine word to use there konii was nitpicking imo

[flint dome]

he'll just dump money into iter or the others

[fast pagoda]

a barrier in the sense of a think you cannot get past

ya

[sage crag]

full equatation is actually E^2 = (mc^2)^2 + (pc)^2

[fast pagoda]

i think the problem is that barrier implies there's something behind it

[sage crag]

where p is momentum

[olive sable]

ye, but i didnt bother to remember that one

[uneven pulsar]

✓

[sage crag]

this is the reason that photons can still push things despite having zero mass

[fast pagoda]

pc = personal computer

[hearty notch]

we've defined the things behind it like tachyons we just don't expect them to exist

[sage crag]

tachyons are poorly defined

[hearty notch]

because they probably don't exist yea

[fast pagoda]

part of the fundamental equations that define our reality

[flint dome]

true... then again we may be still stupid enough to not realise that things may exist beyond it

[fast pagoda]

is

gaming

[fast pagoda]

i certainly am

[sage crag]

ne speed of light is one thing we are pretty confident about

cant go over it without creating infinity somewhere

[uneven pulsar]

Bye I still
Coding dump things on ai

[obsidian mantle]

what about those warps you mentioned

[hearty notch]

gl i

[sage crag]

shrinking a distance and then moving that distance, to circumvent the speed limit

[olive sable]

would that mean, that you can get negative energy for things with negative mass?

[flint dome]

i mean for all purposes our best way of exiting this bitch of a earth's gravity is still strapping millions of tons of explosives to our ass and farting it out at a lunatic manner just to get up there

[sage crag]

in principle energy is a scalar

there is no negative energy

[olive sable]

bwaa

[obsidian mantle]

wait but doesnt distance shrink too when you approach speed of light

[olive sable]

i know antimatter has positive mass, which is kinda sad

its jsut oposite charge or something

scam

[tender river]

E=mc²+AI

[fast pagoda]

[hearty notch]

opposite quark composition but yea

[fast pagoda]

nvidia is writing that down now

[flint dome]

how would one get a negative mass on something

[sage crag]

it appears to shrink but its not actually shrunk

its a relativistic effect

[obsidian mantle]

oh so just visual

[sage crag]

not quite

[fast pagoda]

https://tenor.com/view/ursula-laughing-mirror-little-mermaid-creepy-gif-6083940

[sage crag]

distance as observed by another person would remain the same

but for yourself it would appear to contract yes

[olive sable]

make mass out of nothing, and then you have the negative mass left?
idk im not a scientist, im a sciencefictionwriter at best

[flint dome]

would it increase the mass of something else instead of its own mass

[sage crag]

"negative mass"

[flint dome]

because imagine it in the sense of money
you have + money (no debts, only cash)
or you have - money (you have debts)

[olive sable]

i know it sounds stupid, and it is also stupid, but it would be cool

[obsidian mantle]

does it mean it will appear that you are moving "slower"

to yourself

[fast pagoda]

i have 3 kids and no money but i want no kids and 3 money

[sage crag]

ye as you approach lightspeed you appear to become slower

or more accurately, everything slows around you

in fact, photons dont percieve time at all

[olive sable]

its so hard being a single mom when you dont have children and you're a teenage boy

[sage crag]

which is why its not really accurate to say that they travel

[hearty notch]

everything slows around you yea if you move fast enough the whole universe will age faster than you

[flint dome]

look at yourself in the mirror and cry that the version in your mirror probably has it

[obsidian mantle]

so photon sees its destination right before its eyes but it looks like nothing is moving?

[sage crag]

its more that they "propagate"

[obsidian mantle]

(lets pretend it has eyes)

[fast pagoda]

if you were somehow able to break the speed of light you'd basically just be timetraveling because then you could pretty much send information to yourself in the past

you could be philip j fry

and be your own grandpa

[sage crag]

no its more like, from the photons perspective it already reached its destination

[obsidian mantle]

entire universe is 2d for photon

[fast pagoda]

photon has no concept of time

[obsidian mantle]

right?

[fast pagoda]

yeah that dimension is just gone

[flint dome]

https://tenor.com/view/my-love-story-with-yamada-kun-at-lv999-yamada-lv999-anime-yamada-lv999-akane-akane-confused-confused-anime-gif-16048699582846891922

[sage crag]

universe still 3d

[olive sable]

https://tenor.com/view/rintaro-okabe-anime-steins-gate-gif-3245780294731451925

[sage crag]

time dimension gone though

[olive sable]

steins gate moment

[fast pagoda]

classic banger

[pseudo silo]

Yes

[obsidian mantle]

photon is ann

[fast pagoda]

kek

when spiking light

[pseudo silo]

;0

Technoblade never die

[sage crag]

suddenly universe has ended

[fast pagoda]

hey speaking of discrete vs continuous, there's some debate about whether or not spacetime itself is even continuous or discrete due to qUaNtUm ThEoRy

[olive sable]

erm

[sage crag]

yee already mentioned

[olive sable]

i have no clue

[sage crag]

like i say, no evidence that spacetime is discrete yet

however, lots of fundamental forces are discrete

[fast pagoda]

oh ok i missed that part

[olive sable]

does it even matter tho? we'll never see the end if it is discrete

[pseudo silo]

No

[glad path]

you would be time traveling i think

[sage crag]

if we had a working tested theory of quantum gravity we may be able to identify if spacetime was discrete or not

[olive sable]

my personal guess is its probably discrete, but the only thing i know is i dont know shit

[sage crag]

for now we dont know

[glad path]

actually wouldn't you just instantly die

like

[sage crag]

my personal guess is that spacetime can be infinitely subdivided and is continuous

[glad path]

time is relative, right

[obsidian mantle]

nooo

[glad path]

so

[obsidian mantle]

you have same time perseption for yourself

[pseudo silo]

What

[obsidian mantle]

you will age with same rate

just the universe will freeze

[sage crag]

god im tired of this

[pseudo silo]

Yes

[glad path]

in theory you'd just be going so fast that relative to the outside world, a few billion years would be a single second

[sage crag]

bye bye talk about this among yourself

ne

[glad path]

and you'd just get to the heat death of the universe

[obsidian mantle]

sorry konii i dont have unlimited brain

[pseudo silo]

Not sure

[glad path]

also my theory is probably flawed

anyhow

[olive sable]

i know you know more about the big bang than me, but i thought since space supposedly expanded from there it would have an end?

[silent cloak]

are u talking about the effects of falling into a blackhole

[flint dome]

everyone keeps saying how to go back in time... how to go back.... i ask how the hell you go forward

[olive sable]

oh nvm then

[glad path]

no

that one's easy

death

[silent cloak]

going forward in time is the easy part thanks to time dialation

[pseudo silo]

But if we do

[silent cloak]

gravity effects the 4th dimension

[glad path]

affects*

[fast pagoda]

quick someone bring up the 2 slit experiment where people totally misunderstand what it is trying to demonstrate and then go off on a conspiracy theory about consciousness

oh wait i just did

[silent cloak]

https://tenor.com/view/yummy-tongue-flight-flight-reacts-excited-gif-17057509

[flint dome]

yeah but i dont have a damn blackhole as easy access

[glad path]

quote of all time

[pseudo silo]

No

[glad path]

ngl

[pseudo silo]

[hearty notch]

haha this guy can't pull black holes

[fast pagoda]

put it down in the louvre to replace that lost stuff

[obsidian mantle]

observing quantum particle changes its behaviour

[olive sable]

from what i remember it just makes a gradient where the waves collapse or increase each-others' amplitude. something about probability idk

[pseudo silo]

🧉

[obsidian mantle]

and by "observing" we mean bombing it with other particles

[flint dome]

i cant even pull women how do i pull a blackhole bruh 💀

[pseudo silo]

Yes

[obsidian mantle]

how can this happen

[silent cloak]

hologram theory

[flint dome]

just put a small camera near it smh why disturb it with other particles

[pseudo silo]

No

[obsidian mantle]

doesnt work, too small

[glad path]

observing an electron fixes it to one place, otherwise it's in all places at once

[obsidian mantle]

transparent to any light

[glad path]

yeah and observing means like

[hearty notch]

there was a good veritasium on the two slit experiment and all its implications can't remember the name tho

[glad path]

tryna see where it actually is

[olive sable]

you need a microwave, a 42 inch crt tv, a mobile phone, a dr pepper, and a cool lab coat

[glad path]

not just looking at it i dont think

[flint dome]

"nah bro if its transparent just put night vision on it then you can defo see it"

[pseudo silo]

Yes we do

[hearty notch]

observation in a quantum sense is just literally doing anything that forces the wave function collapse

[obsidian mantle]

konii left, this conversation is doomed

[olive sable]

[fast pagoda]

people just misunderstand what observation means and think it has to be a conscious person doing the "observing" and that somehow us looking at something is some sort of magic effect that causes the info to manifest

[obsidian mantle]

nobody to resolve the questions

[pseudo silo]

Fuck that Doomsday

[flint dome]

then switch to a topic we know

[pseudo silo]

Ok

[obsidian mantle]

which name do i give to my chess bot

[hearty notch]

steve

[flint dome]

ligma master

[fast pagoda]

chess_bot.py

[glad path]

why did ijust fail to load my own bio

[silent cloak]

chesster

[obsidian mantle]

do i make super cringe name

[stray dragon]

shitass

[glad path]

what

[maiden geyser]

when will phone manufacturers let 685 die

[fast pagoda]

FUCK

chess

[glad path]

but that's me

[fast pagoda]

holy shit

[flint dome]

t i have a question

[silent cloak]

thats what my roomates nickname is uni was

well it was "ass shit"

[hearty notch]

wow why would you block urself like that

[silent cloak]

but same thing

[stray dragon]

did he deserve it

[fast pagoda]

my boy dingleberry

[nocturne olive]

Oh yeah the Chess competition
I forgor
Guess I won't have a bot

[silent cloak]

he was from italy and one time he said "im hungry as shit" but his accident made it sound like ass shit

[nocturne olive]

Sad

[stray dragon]

yeah same

[silent cloak]

so we started calling him ass shit

[stray dragon]

woops!

[fast pagoda]

me just now

[pseudo silo]

Ok

[fast pagoda]

mmm ass shit

[nocturne olive]

Shir disappeared from here so everyone forgor

[flint dome]

our nicknames for each other are "street light pole", "dipshit", "horse shit", "glue sniffer"

[fast pagoda]

classic italian dish

[silent cloak]

i forgot what they called me

[hearty notch]

for some reason graduation mans name existing on my screen got rolling girl stuck in my head idk how that works

[pseudo silo]

Cat

[fast pagoda]

lmfao

[stray dragon]

which one are you

[flint dome]

that smug face of garfield makes it better

[fast pagoda]

i ham happy this is my pfp for this

[glad path]

it does

unfortunately

nothing here will ever really hit starboard

because

[flint dome]

my name is squirrel fucker

[fast pagoda]

there's 6 of us

so

yea

[glad path]

there's not enough people for it

[fast pagoda]

or w/e it is

[glad path]

they gotta lower starboard req here to 5

[pseudo silo]

:?

[fast pagoda]

they have starboard tuned for gen and nn which can get like 60 stars in 2 seconds

[stray dragon]

interesting.

[fast pagoda]

s-sandy cheeks?

[olive sable]

we did get the triangle art to hit starboard that one time

[stray dragon]

very interesting.

[glad path]

did we

[olive sable]

yes

[pseudo silo]

👍🏿

[glad path]

that looks like its still loading

looks greyed out

[pseudo silo]

My bad

[glad path]

i forgot toned emojis exist lol

[flint dome]

came from a misconception when i was in a park and a squirrel kept pissing me off and biting me so i kicked it

then i accidentally left the mic on while flushing and the kids thought i flushed the squirrel and some kid said "he probably did that after fucking it"

the rest is history

[glad path]

what

why did you have a mic

huh

[fast pagoda]

stupid fuckin mistakes man

[stray dragon]

what kids

[flint dome]

all this time i was on a discord call because we were doing an important project

[glad path]

ok

[stray dragon]

i see

[glad path]

this is getting more concerning

[pseudo silo]

No

[hearty notch]

my first thought was irl streaming lmfao

[flint dome]

school grade 12 final project

[glad path]

https://tenor.com/view/scary-jumpscare-november1st-november-mariah-carey-gif-23694445

[pseudo silo]

No

[flint dome]

one of them is just straight up called senator gaylord nelson

[glad path]

what

[flint dome]

(hes gay allegedly)

[hearty notch]

gaylord is a normal name believe it or not

[olive sable]

i cant find it rn bwaa

[fast pagoda]

i have a video of me in my sunday best (one(1) pair of boxer briefs) getting convinced to dive into a lake when it was like -15c outside, then running away into my car

[hearty notch]

hell yea

[fast pagoda]

my hometown is unironically called gaylord

[glad path]

ok??

[flint dome]

i know its the guy who made world enviroment day

[hearty notch]

lol

[fast pagoda]

i was meaning to reply to jester's anecdote but then i saw gaylord lol

[hollow matrix]

tf did I just open discord to

[flint dome]

must have been chilly

[fast pagoda]

brainrot

[fast pagoda]

it was

idk why what you said reminded me of that

just stupid ass shit back in the day i guess

[olive sable]

found it

peak programming chat performance

[flint dome]

no we do stupid shit all the time

its important to do dumb shit

[hexed grove]

😭

[hearty notch]

it won't let me add another star this is 1984

[olive sable]

its such a shitty drawign compared to everything else on starboard, its beautifull

[flint dome]

someone advertise it in genchat they will defo star it

[hearty notch]

what's the lore behind red triangle

[olive sable]

vulkan

[fast pagoda]

i am the subject of a halloween legend at michigan state, at least for a while, because i was gumby wearing a lebron james jersey for halloween and got so crunk that i blacked out and woke up without the costume

and found out the next day that, the previous evening during the time i lost, there was a kerfluffle involving several campus police tackling lebron gumby to the ground only for the wily man in the suit slithering out of it and escaping into the woods

[hexed grove]

the vulkan triangle

[olive sable]

we did a contest to see who could render it the fastest

[fast pagoda]

making a triangle render REALLY FAST

[hearty notch]

I see

[olive sable]

quack won at 102000fps

[hearty notch]

lmao ty

[olive sable]

my pc topped out a 44K

[hearty notch]

p2w...

[olive sable]

yep

[hexed grove]

yall should i learn vulkan

[fast pagoda]

i got like 68-70k i think

on nobara linux baybeee

[hexed grove]

i know opengl

[flint dome]

how the fuck did you tapeworm your way out of being tackled in a gumby suit

.... did you wake up buck naked the next day?????

[hexed grove]

but its a bit aging

[olive sable]

are you a masochist? -> yes -> go ahead
-> no -> maybe don't

[fast pagoda]

nah i had clothes on underneath i just didnt have the suit anymore

nor did i have my damn gold chain necklace

[flint dome]

rest in police custody, gold chain necklace

[hexed grove]

okay

will do so

[olive sable]

[amber fractal]

I was at 47K fps for triangle, Still beating Sam's PC on laptop

[hexed grove]

will it give me the last 2 frames per second on my rotating cube

[fast pagoda]

finally found the repo

worst name world

[olive sable]

igpu is where its at for low cost high frequency renderinng

[hearty notch]

lol

[fast pagoda]

this was the triangle leaderboard