the NVIDIA RTX 5090 is here: $1999 MSRP
| Deep talented shitlib | 01/06/25 | | Concupiscible irate lay | 01/06/25 | | garnet puppy | 01/06/25 | | Deep talented shitlib | 01/06/25 | | garnet puppy | 01/06/25 | | Deep talented shitlib | 01/06/25 | | Multi-colored den goal in life | 01/06/25 | | Alcoholic menage depressive | 01/08/25 | | Transparent newt | 01/06/25 | | disturbing wagecucks | 01/06/25 | | disturbing wagecucks | 01/06/25 | | Multi-colored den goal in life | 01/06/25 | | Multi-colored den goal in life | 01/06/25 | | Deep talented shitlib | 01/06/25 | | Multi-colored den goal in life | 01/06/25 | | Deep talented shitlib | 01/06/25 | | disturbing wagecucks | 01/07/25 |
Poast new message in this thread
Date: January 6th, 2025 10:36 PM Author: Concupiscible irate lay
what do you care
just charge it to one of your credit cards and blow off the bill
(http://www.autoadmit.com/thread.php?thread_id=5660081&forum_id=2#48525950) |
 |
Date: January 6th, 2025 11:05 PM Author: Multi-colored den goal in life
The only reason I bought a 4070 was because of the low power draw and the fact that it will work fine in older pcie slots. This allows me to run it as either a primary or secondary GPU and pass it through to a container running local LLMs through a web ui. It doesn't really matter if I use it as primary since the system is headless anyway.
Secondary is a cheap af Intel GPU with all kinds of hardware encoding, and that be dedicated to another container running handbrake. If ur using an Intel CPU with integrated graphics, you actually have a third GPU you can use to run a monitor in a pinch.
It sucks ass that I'm limited to 12gb, but I'm finna fix that real soon by switching to Apple. In the meantime, any Nvidia GPU with more than 12gb is still going to be an energy HOG. This series of GPUs only makes that problem worse, if anything.
(http://www.autoadmit.com/thread.php?thread_id=5660081&forum_id=2#48526002) |
|
|