The localllama/local LLM community ridicules AMD basically every single day.
They have the hardware. They have 90% of the software. Then they waste it with absolutely nonsensical business decisions, like they are actively trying to avoid the market.
2 phone calls from Lisa Su (one to OEMs lifting VRAM restriction, another to engineers yelling “someone fix these random bugs with flash attention and torchetune , now,”) would absolutely revolutionize the AI space, just to start… and apparently they couldn’t care less. It’s mind boggling.
The localllama/local LLM community ridicules AMD basically every single day.
They have the hardware. They have 90% of the software. Then they waste it with absolutely nonsensical business decisions, like they are actively trying to avoid the market.
2 phone calls from Lisa Su (one to OEMs lifting VRAM restriction, another to engineers yelling “someone fix these random bugs with flash attention and torchetune , now,”) would absolutely revolutionize the AI space, just to start… and apparently they couldn’t care less. It’s mind boggling.