Python packaging is definitely in need of something, but my sense is that "yet another tool trying to be a catch-all package/environment management interface" is not it.
I finished this recently, and ended up with a mixed-but-overall-negative view.
As someone who had a reasonable aptitude for mathematics in the past but essentially no practice in the last 10 years, it seemed like the content varied wildly in how well it built on previous chapters, or more generally adhered to the "introduction" label. The programming-based content was thin to the point of being pretty much irrelevant, leaving the whole as something more like a scatter-brained textbook for an odd curriculum rather than a refocused version of a sensible one.
Fully agree, but the announcement is pretty bare-bones so no reason to think that adding InfiniBand is totally out of the question, particularly since this offering seems to be pretty far from a standard GCP experience.
Having experimented with IB on Azure, I can comfortably say that the problem is not so much the fabric but the general usability of Azure itself - hot garbage doesn't do it justice.
They don't call it out in the announcement, but this was published in the same week as the "Supercomputing" conference - GCP hasn't had an answer for HPC workloads thus far, so maybe this is their attempt.
If that's the case, can you describe how you ended up with customers?
Whenever I consider entrepreneurship as an option for myself, it is always the customer-facing sales part which leaves me feeling like I would struggle.
They have been describing this as shipping (in the Dell systems mentioned) for a while, I would guess 2 years. It isn't clear to me where the hardware is going or why the company hasn't made more of a splash if performance is so good.
Coming from the scientific computing realm, I've only ever done "real" code in Python until a couple of weeks ago when I was forced to do some work in Typescript. Once I got over the initial worries about the best method to install npm to avoid Python-esque problems and gave nvm a try, I was very pleasantly surprised by the package management process and was able to just get on with the work.
I've tried various Python env management solutions in the past (mostly leaning towards conda), but had recently settled on just using separate LXC/LXD containers for each project.