When do we get to break away from binary as the core of computing?
When do we get to break away from binary as the core of computing?
Other urls found in this thread:
news.ycombinator.com
twitter.com
Nice try SJW you can't push your quantum genders on me
once they publicize quantum computers
so, never
im sure they'll come out with it after graphene's used up its time
At the same time that the rest of humanity breaks away from base 10.
When we try the ternary system again, though I'd wonder how many things going from 0, 1 to -1, 0, 1 fuck up
kek
Interestingly I just saw a post about analog computing on hackermemes
news.ycombinator.com
Also I think hardware neural network circuits won't necessarily have to be binary, it wouldn't be too much of a stretch to implement spiking neurons and shit in hardware I'd say.
How would if statements work if theres a third state?
maybe
Just like switch statements obviously.
But there are other possibliities, especially when you have way more than 3 states, like fuzzy logic stuff or probabilistic computing.
Quantum in 20 uears
Can you repeat the question?
Can't wait for my processor to fuck up when someone decides to use microwave.
Or booleans for that matter
>Inb4
false, somewhat, true
Do I look like the guy in the picture?
no, but you look like a german faggot
Do you vape?
No, but I am computer addicted
Yes
No
Both/Neither
Mu
When we get to biological computing, which will be much more likely than quantom computing.
...
>throw away cracked graphene panel on the ground
>never stops
>creates a new equator from here to china
>water never passes through betwen two half spheres
shit's dangerous yo
>uears
how many years is one uear? or is this a unit for ears?
4 ears make a uear
>never stops
wat
>Also I think hardware neural network circuits won't necessarily have to be binary, it wouldn't be too much of a stretch to implement spiking neurons and shit in hardware I'd say.
spiking neurons are shit and much of the research into making neural networks efficient works, ironically, with binarized neural nets.
there are efforts to create huge spiking neural nets (derided by most of the machine learning community as an useless endeavor, considering how little we know about them we can't possibly hope to implement them right now)
neural network hardware also already exists, check out TPUs from google. One of the main selling points is that they can work with 8bit quantization very fast, compared to fp16 which is the least amount of precision you can get on today's GPUs. Current architectures don't need that much precision to perform at their best, quite the opposite.
john von neumann was designing it before he died. was a genius that guy...
I imagine that switch statements would probably resolve to default rather than a definite yes or no
switch(variable){
case 1:Print("1"); break;
case 2:Print("2"); break;
default:
Print("default);
}
Trinary is more than possible now...
It is the most practical known way of computing. Period. Quantum computing isn't going to be a replacement, it's simply a value add. Linear calculations can't be done in a quantum computer, and software as it exists now, is linear.