57 more posts in this thread. [Missing image file: ]
Hi, I'm a market researcher for a certain large company, and I'm here to ask for your opinion about a project that's in the works.
Basically, it's supposed to be a powerful all-in-one computer like an iMac, except that it's not, and it would come in three different tiers.
Here are some rough specifications:
>All models will have a stylish aluminum casing, and a wireless combo is going to be shipped with them
Intel Core i5-4310M
GeForce GTX860M 2GB
60GB SSD + 1TB HDD
27" 2560x1600 MVA
Intel Core i7-4712MQ
GeForce GTX970M 6GB
120GB SSD + 1TB HDD
27" 4K 3840x2400 MVA
Intel Core i7-4910MQ
2x GeForce GTX980M 8GB
250GB SSD + 2TB HDD
28" (smaller bezels) 5K 5120x3200 MVA
I don't know the final prices, but they're likely going to be around $1200 for the base model and under $4000 for the top one.
36 more posts in this thread. [Missing image file: ]
My early 2011 Macbook pro is about to break, partially because it's made by Apple and /g/ hates it, and partially because of a GPU failure and the fact it's taking effort to start up these days. Applecare has just run out, and they told me it'll be about $1200 to repair, because they are faggots.
So, I plan to piss you all the fuck off and buy a new macbook pro. I'm thinking I'll go for either the $1800 13-inch or the $2000 15-inch, either of which I can get for $200 off seeing as I work at a local University.
What do you think, /g/?
9 more posts in this thread. [Missing image file: ]
How to build sub-supercomputers at home for artificial neural networks? What kind of power has to be increased in computers in order for them to perform better at the calculations associated with huge artificial neural networks? Is it all about the CPU? Or is RAM important as well? I know that GPU isn't.
What should one shop for if one is interested in very powerful processing power computers with very low GPU power, and a very small/no hard drive? Are there ways to build a computer without HD and with minimal GPU (which would just handle Gentoo), just with powerful CPUs (and RAM if needed)?
I'm not some retarded teenager or a newfag - I'm a first year math graduate interested in applying artificial intelligence techniques for proving theorems as well as constructing new mathematical definitions, which would then be selected by some kind of an evolutionary algorithm with at least minimal human input.
I tried using Python's artificial neural network library for this, but it takes a very long while to prove even most basic and obvious theorems which take few steps to prove (I have a 2-core ~2400 mhz (?) CPU and 4GB of RAM). I figure I could prove complex theorems with a very large neural network and more computational power.
How many flops can I expect to get with a budget of $1000, assuming I already have a screen, mouse, keyboard, etc.?