Wouldn't this mean someone with sufficient hardware could lift the SOTA model weights off the chip? Or are you saying that these chips would only be used internally by these companies and not sold to the public?

▲

dumberquestions 2 hours ago | parent | next [-]

I wouldn't expect companies not sharing their weights today to be any more likely to share them if they're on hardware, this doesn't sufficiently hide weights from a local user.

▲

snek_case 2 hours ago | parent | prev | next [-]

The weights are very unlikely to be on the chip itself. That wouldn't work for SOTA models that are terabyte scale, even quantized. This is probably an accelerator for specific kernels in the model, but the weights are likely loaded from memory. The chip may have SRAM to store some of the weights temporarily during inference.

	▲	foltik an hour ago \| parent [-]
		At least in the case of Taalas the weights are physically encoded directly on the chip. It’s composed of 4-bit multiplier cells that compute all 16 possible results in parallel. The top metal wiring layer physically selects the one that corresponds to a multiplication with that cell’s constant weight, and routes it to the next layer.

▲

syntaxing 2 hours ago | parent | prev | next [-]

I don’t get why this is an issue? You can run Claude/OpenAI SOTA models through Amazon bedrock. These weights have to live somewhere to run on Bedrock.

	▲	wmf 31 minutes ago \| parent [-]
		somewhere = an AWS data center with multiple layers of security and NDAs They won't sell/rent/license the weights to an end user at any price because they don't trust your security.

▲

amazingamazing 2 hours ago | parent | prev [-]

One idea would be to use an open model.