← Home ← Back to /g/

Thread 106571886

10 posts 6 images /g/
Anonymous No.106571886 >>106571912 >>106571943 >>106572039 >>106572483 >>106572760
Do you see a future in photonics based computing? Lot of Silicon Valley ventures are putting serious money behind this company for some reason.

https://lightmatter.co/products/envise/
Anonymous No.106571912 >>106571954
>>106571886 (OP)
And the most important thing: ITS CHEAPPER
Anonymous No.106571942 >>106571954
if it were possible it would already exist
Anonymous No.106571943 >>106571954
>>106571886 (OP)
It's probably just jews xeroxing elon or the zuckerberg looking girl
Anonymous No.106571954
>>106571943
>>106571942
>>106571912
Not knowledgeable on Silicon Valley start ups but why do so many of positions seem redundant or totally useless?
Anonymous No.106572034
seems retarded
inb4
>LIGHTMATTER PLATFORMS REVEALED TO BE 500 DATACENTERS IN INDIA
Anonymous No.106572039
>>106571886 (OP)
its much more realistic than quantum computing but so far scaling it has been a problem. Magnonics is a similar technique with the same frequency based computing advantage. Or flux devices thats the real end game.
Anonymous No.106572047
photons are just light nigga
Anonymous No.106572483 >>106572760
>>106571886 (OP)
The reason is low power consumption. I've also seen it's possible to cool parts of a chip with laser, it'll probably be integrated eventually to boost clock speeds in critical paths.
Anonymous No.106572760
>>106572483
> I've also seen it's possible to cool parts of a chip with laser

Lies.

Lasers are not a chip cooling devising, remote possibility that some data transfer or types of compute can be thermally cheaper using light but that's not the same as a laser cooling a fucking chip, its just it being more efficient.

If you thinking of laser cooling, that is a thing in some niche physics applications that have absolutely nothing to do with cooling a fucking computer chip you fucking brainlet.


>>106571886 (OP)
Mostly I am interested in free space diffraction based computing for solving problems with many candidate inputs and few valid solutions. This silicon photonics stuff I would only be interested in using it conjoined with aforementioned types of things only to not have to domain switch between optical and electrical so much.