It’s still a fantastic computer.

I use it as a server and it’s very very fast, especially threaded workflows, and IO is fast.

Just don’t buy it expecting to replace paid AI services. And don’t buy it for AI dev, on paper it should be good, but driver issues. DGX Spark is better if yo want an AI dev machine.

This shit is awful.

You end up with managers hiring bad candidates to save the rest of their team.

When they had their schism over Altman a couple years ago safety died.

Can we do like a Canada plus Europe friendship exchange?

I definitely buy Canadian and European first, then anywhere else, then US now.

That’s abuse and harassment.

There’s no other way to look at it. It is not that relative, it’s fraud to claim it is, and it’s just disgusting to pull at the heart strings like this to make money.

I can’t tell if he’s going to come out and take credit for this as he tricked Europe into funding their military and he’s a genius.

meanwhile the cost to US the soft power and economy is going to be immeasurable and last for at least an entire generation.

Wow, that’s shameful and disgusting.

Thanks for your detailed correction

I did not able that last part

So the modified handgun would be a rifle with sawn off barrel, I can’t recall the length but anything before a limit of automatically considered a handgun by the RCMP.

It’ll be interesting to see if they discover where it came from. I’m still suspicious of the PAL plus removed weapons. Is it possible they bought a gun under the table after their weapons were confiscated?

How did they get this to run, I thought browser stuff was capped to 4gb?

LOL

ROCm has got to be the worst software I’ve ever dealt with.

It would not be worth it as a replacement for Claude.

80% of my issue is that it’s AMD and their drivers are still awful. 20% is that the token generation speed very slow, especially compared to commercial models running on dedicated hardware. MOE models are fine, dense models are too slow for meaningful workflows. ComfyUI is decent, but I’m not seriously into image gen.

I have a lot of fun with it, but I have not been able to use it for any actual AI dev.

Yes

And they could have just started with a labour skill visa.

They’d clean up illegals because they’d become legal, and they’d probably vote for tax cuts.

Do you really think there’s a 50 foot tall transparent cow out there?

There’s no fall through case, it’s ambiguous what the room state should be

High RAM for MOE models, high VRAM for dense models, and the highest GPU memory bandwidth you can get.

For stable diffusion models (comfyui), you want high VRAM and bandwidth. Diffusion is a GPU heavy and memory intensive operation.

Software/driver support is very important for diffusion models and comfy UI, so your best experience will be Nvidia cards.

I think realistically you need 80gb+ of RAM for things like qwen image quants (40 for model, 20-40 for LORA adapters in ComfyUI to get output).

I run an 128gb AMD AI 395+ Max rig, qwen image takes 5-20 minutes per 720p qwen image result in ComfyUI. Batching offers an improvement, reducing iterations during prototyping makes a huge difference. I have not tested since the fall though, and the newer models are more efficient.

It did bother me there was no fall through.

I think it’s time to admit that “elite” institutions aren’t really better than any others, they just power a nepotism machine and claim the gains as then being better.

Some schools have better networks of teachers and researchers, but this is a con.

The rest of us have to follow the rules while the rich get into the best (reputation) schools, are protected from pressure, and get to pass.

Just look at that Wharton prof who said Trump was the worst student he ever had. Trump got his degree, why didn’t he get flunked if he was so dumb.