as long as it’s their hardware

This is the problem for me. AMD exist, Mac exists, Intel exists, and they rely on llama.cpp which now has the devs getting their paychecks from Nvidia… They’ll just “direct primary focus” onto cuda development and oops we didn’t touch SYCL support for 8 months… Whoopsiedoodle.

I literally just bought a b70… Fucking Nvidia…

Yes I know its open source, but the devs currently have a decent pace with updates. Relying on unpaid devs to care about SYLC when most people don’t run it anyway is probably worse than relying on Nvidia paid devs to eventually get to it :/

“together we will regulate model access, slow AMD, Mac, and Intel support on llama.cpp, and generally enshittify the experience like all corporations do! Welcome to the future, the same future as every other corporate acquisition!”

Like, I’m playing Bully. Render the locked 30, barely consume any power and shut up

I am getting older and honestly only recently came to appreciate 60fps, it blows my mind to hear people say shit like “doesn’t do 120fps, unplayable.” …like wtf?

We know everything is more expensive, but have you tried just learning less?

…

Also for those after your 30s, strengthen your core. I used to have back pain but no longer!

A 70 year old man got me to work out. He just does 10 mins a day, 3 times a week. If I can’t do that wtf am I doing?

One 80 year old destroying our future, and Death destroying our past. What a shit year

“Don’t worry buddy, we’re trying as hard as we can to lose. Your families grifting won’t come to an end on my watch.”

Because they know they’ve been lying about climate change to keep feeding the oil oligarchy and now they need to protect the other oligarchs from the repercussions of their previous actions.

I think that no matter what happens we’re going to foot the bill. There’s no way in hell these assholes will ever face consequences of any sort. They already destroyed consumer electronics markets, and will bring the rest of the economy down with them if they fail, but they themselves and the parasite investors will never lose. We’ll have to pay…

Thanks for the info! I believe I’m running MTP n-max 4 if I’m not mistaken which seems to be holding out well

At first I was mislead by some faulty benchmarking I was doing and thought that llama.cpp + SYCL was worse than Vulkan stock. Vulkan stock couldn’t run past 10t/s with MTP for some reason, so after a whole bunch of shenanigans I ended up again back on llama.cpp + SYCL but was lead to believe MTP was hurting me so I didn’t retry until last night. I finally got the 35-40t/s with llama.cpp + SYCL and MTP and 800PP with AOT over using JIT. I drop to 25ish after 40-50k context and stay kinda flat 12ish at 160k

I didn’t want to move to Linux to try vLLM so this was all just trying to fight windows b.s.

If you’re on windows and fighting with the VRAM offload after 70 seconds like I was: HKLM\SYSTEM\CurrentControlSet\Control\GraphicDrivers

EnableRuntimePowerManagement set to 0

as well as a 500 dollar GPU

(Cries in $1,300 Intel arc fighting tooth and nail to break 25t/s…)

I’m not seeking to justify their actions, but explain why the change seems to have happened around that time. I’m not convinced it’s totally on gen X to have that sudden change, but more that the entire population was affected by the explosion of propaganda

I am Jacks complete lack of surprise.

:/

Other things that have changed in the same time frame is the explosive growth of Reich wing broadcasting. Rush Limbaugh and the like coming to reach millions of people and then the birth of fox news. That has made things much worse in the last 40ish years

Narrator: they did in fact take it.

It has been some time since my initial comment so at the time I was mainly using LM studio. Qwen 3.6 a3b is the MOE and it does work well on my card, but the dense model that is more intelligent/capable is the Qwen 3.6 27b which doesn’t fit on the card and does get offloaded, but offloading cuts the speed down to like 1/tps.

I have since found a version of the 27b model that is “quantized,” for lack of a better term, differently and has to be run through TabbyAPI which gets back to 30ish tps. It can’t offload so it must fit fully on the card which keeps the speed high. Might be worth a look if you’re interested, the only downside is that with my 16gb card the context limit has to be kept pretty low ~40k if I remember correctly

But guys he doesn’t take his presidential salary of 400k/YEAR so obviously hes incredibly generous and not profiting off the presidency!

…/wrist

This is some of the dumbest shit ever…

Republicans: I have concerns but I’ll vote for xyz.

6 months later: I’m very disappointed in xyz, we definitely couldn’t see this coming.

Over and over and over and over and over and over and over and over and over and over and over and over and over… It’s like the “say the line Bart!” Meme…

It’s 100000% guaranteed he revives this taxpayer giveaway to the absolute worst scum. It’s insanity that anyone actually believes when Republicans say they got confirmation that the fund is dead.

Motherfucker is going to do an 18/22!