[D] Any recent ML server build guides or recommendations?

I see build guides for nvidia RTX 2000, 3000 series, but none for 4000-series. Unfortunately the knowledge shared has aged rather quickly as GPUs have gotten fatter, longer, taller.

Has anyone managed to fit multiple rtx4090 in a single ML server? If so, any suggestions?

Thank you

aged build gpus guides knowledge machinelearning managed multiple nvidia nvidia rtx recommendations rtx series server suggestions