joennlae

agreed. It is not easy to benchmark these models. It just the date of index changes have some „spurious“ correlations

Interesting that changes in the index always happen when „non us big tech“ models seem to be a little to good for comfort.

„The joint service offering will provide the bundled compute, model, and licenses, while ensuring customer-specific design data is protected.“

I am not sure if Nvidia want to send their chip designs to OpenAI.

EU AI Act is exactly that. Mandatory „stop button“ for High Risk Systems.

what a find

How can I make sure that each github runner uses exactly one cpu core?

+1. Does someone know how to do that?

Author here: Let me try to give an overview as I saw some questions repeating itself.

* This accelerator is for an Edge/Inference case, so there is no training on this chip.

* We introduce a differentiable form of Maddness, allowing Maddness to be used in e2e training and present an application -> ResNet.

* We are still in the process of understanding how this will translate to transformers.

* The goal was to show that Maddness is feasible with a good codesign of the hardware.

* Compared to other extreme quantisation (BNN/TNN) and pruning schemes, this is more general as it replaces the matmul with an approximate matmul.

* The model architecture is not fixed in hardware. It is „just“ a matmul unit.

I hope this helps :-)

Thank you for the feedback :-)

We have to be careful with the comparisons we make. The TPUv3 is a training and datacenter chip and not an Edge/Inference chip. They optimise for a different tradeoff, so while the comparison looks good, it is unfair.