Hasan Genc

`gemmini_config_ld` is just [`config_mvin`](https://github.com/ucb-bar/gemmini#config_mvin-configures-the-load-pipeline) and `gemmini_config_st` is just [`config_mvout`](`https://github.com/ucb-bar/gemmini#config_mvout-configures-the-store-pipeline`). (Sorry about the inconsistent naming). > Whether the setting of the address has a special meaning or a special purpose. The...

> I think this example is just transferring A to C, and it should not perform the addition operation, a bit I didn't understand. As we describe [here](https://github.com/ucb-bar/gemmini#memory-addressing-scheme), if the...

> I don't understand the role of `preload`,`preload_zeros` The `preload` commands are used to preload either a bias or weights into the spatial array. These preloaded values will remain stationary...

That code is just trying to determine maximum tiling parameters based on the scratchpad size in Gemmini. It assumes that half the scratchpad will be allocated for matrix A, the...

> On an exciting note, I used MLIR to do some support for the Gemmini software part and am now trying to run a model of ftlite. That's great! I'm...

Hmm, supporting newer formats, as you noticed, can be difficult. However, if the _innermost dimension_ is the same, then changing the format is only a matter of changing the DRAM/L3/L2...

Haha, why don't you email me directly at hngenc [at] berkeley [dot] edu, and I can give you my personal thoughts about grad school vs industry

Sorry, it looks like newer versions of Chipyard no longer generate the old `top.v` file. However, in `gen-collateral/`, you should be able to find `ChipTop.sv` and `Gemmni.sv`, which might be...

The spatial array in Gemmini is double-buffered; you can load in new weights into one set of registers, while continuing to perform matmuls with the old weights in the other...

> I was wondering if it would be feasible to also add a floating point implementation of Softmax and GELU? That would be a great feature to have for sure,...