Tom Lin

Related: https://github.com/UoB-HPC/BabelStream/issues/123

We probably need to update output padding as well, A100's bandwidth is getting close to breaking the formatting.

@tomdeakin I wonder whether we should template/macro the induction or not and make it a runtime option (e.g `--counter=uint32_t|uint64_t|intptr_t` ). It's quite possible changing the loop induction's type will alter...

See https://github.com/UoB-HPC/cloverleaf_tbb/blob/23053cfc3d702bf289a07437577672f7f8fe7a9c/CMakeLists.txt#L9-L14 which integrates FetchContent TBB with CloverLeaf

I'll check this again with the latest patches.

> @rob64j @tom91136 Are you okay with being included as maintainers on this package? Yes

Ok, I've check the source of Etlas, it seems that [HttpUtils.hs](https://github.com/typelead/etlas/blob/6636e02fbe8719070cc52f3075ba88939e5b7b81/etlas/Distribution/Client/HttpUtils.hs#L320)'s implementation of curlTransport behaved strangely with [Gow](https://github.com/bmatzelle/gow). I've confirmed this by removing Gow from my PATH; adding `--http-transport=powershell` also...

Hmm, there isn't a failure to speak of, etlas seems to forget about the curl process entirely and think that the index file is not there. Here's the output with...

I'll compare Gow's curl with a generic linux distro's curl to see if the behavior is different. I think this is becoming more of a system configuration issue and has...

I've got a SYCL2020 USM version implemented in https://github.com/UoB-HPC/BabelStream/commit/3f7bb631e18dd80b7798c7bb8218b041197b1975, will try to merge at some point.