Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Ask HN: What startup/technology is on your 'to watch' list?

news.ycombinator.com

Post reply on HN
Print view View original

181–190 of 257 posts

Re: Ask HN: What startup/technology is on your 'to watch' list?

#181

Post by rewq4321 »

rklaehn wrote: post #123

https://www.unisonweb.org/ Functional programming language where the canonical representation is a content-addressed directed acyclic graph. Solves all kinds of problems from dependencies to deployment and moving code between nodes in a principled way. The language itself is inspired by haskell, but has a principled and clean solution to the coloured function problem of async programming, and a simpler way to compose…

Interesting! Deno's official module registry stores immutable tagged versions of open-source projects, but there's a project called Nest.land[1] that takes the immutability one step further by putting modules on the arweave[2] blockchain.

I'm not sure where the JS standardization process is at with import integrity checks for non-script-tag imports, but Deno has lock files and integrity checking built in.[3]

I like the idea of a language that's built with content addressing from the ground up. I dream of being able to import IPFS urls (or something like that) directly within JS. Although that wouldn't be as good as a language that forced usage of content-addressed imports, since that way you don't have to scour the code for any sneaky dynamic imports (especially since in JavaScript there's `eval` and the like, and they can be obscured).

I guess this is partially solved by Deno's ability to limit the network requests to specific domains like:

    deno run foo.js --allow-net=deno.land,foo.com
I'll definitely be keeping an eye on this project - thanks for sharing!

[0] https://deno.land/x

[1] https://nest.land/

[2] https://www.arweave.org/

[3] https://deno.land/manual/linking_to_external_code/integrity_...

Re: Ask HN: What startup/technology is on your 'to watch' list?

#182

Post by creztoe »

neeeeees wrote: post #66
hurflmurfl wrote: post #62

Earlier quoted context omitted.

Could you please expand a bit on this thought? I personally haven't used almost any API automation tools yet, although I do keep on looking in that direction from time to time. It would be great to hear another perspective.

I think I can answer part of this. I worked on building GraphQL APIs dynamically from Rails models as sort of a pet-project at Shopify a few years ago. It’s not a difficult task, but when discussing viability with the APIs team, the general consensus was that you don’t want your API coupled to your data. Most of the time, you don’t want to expose everything. Of course, you can start describing what to omit, but now y…

That sounds awesome and also a huge headache. I've found its much easier to work with an ORM that is specifically designed for GraphQL, like TypeORM[1]. Otherwise you are just kind of trying to force a square peg in round hole (like Graphene[2] for Django). As for not exposing some data, just use "private" schema directives, etc.

[1] https://github.com/typeorm/typeorm [2] https://github.com/graphql-python/graphene

Re: Ask HN: What startup/technology is on your 'to watch' list?

#185

Post by rewq4321 »

37ef_ced3 wrote: post #178

AVX-512 neural net inference on inexpensive, CPU-only cloud compute instances. GPU cloud compute is almost unbelievably expensive. Even Linode charges $1000 per month, or $1.50 per hour (look at the GPU plans: https://www.linode.com/pricing/#row--compute ) An AVX-512 Skylake-X cloud compute instance costs $10 per CPU-core per month at Vultr ( https://www.vultr.com/products/cloud-compute/ ), and you can do about 18 De…

How could it be competitive with GPUs in terms of price-per-unit-performance? Seems like GPUs are expensive only because you can't rent a small portion of a GPU? But shouldn't that be possible with GPU virtualisation?

Or is it the case that if you virtualised a GPU up into tiny pieces, the memory-to-flops ratio would be way off what's needed for inference? Or the virtualisation overhead would be too big?

Those are all genuine questions, just to be clear - this is not my area of expertise.

Re: Ask HN: What startup/technology is on your 'to watch' list?

#186

Post by dyeje »

progval wrote: post #166
balfirevic wrote: post #158

Earlier quoted context omitted.

It's a standard way to make pages dynamic now. AJAX just means making http requests from javascript instead of reloading the full page.

I meant using AJAX to transfer XML that gets inserted directly into the DOM. Of course that was never the only use of AJAX, but it used to be much more popular than it is now, and it's what Hotwire is trying to make popular again.

It's not quite the same though, there's a light framework involved for defining how the html is inserted, it's over websockets, new request types, etc.

But point taken, I recall using similar techniques a long time ago. I think the pendulum is swinging because front-end development has gone down such a rabbit hole with state management that people are wondering if they can remove all the overhead of redux, translating JSON, client side routing, etc. In most cases, they probably can I reckon.

Re: Ask HN: What startup/technology is on your 'to watch' list?

#188

Post by 37ef_ced3 »

rewq4321 wrote: post #185
37ef_ced3 wrote: post #178

AVX-512 neural net inference on inexpensive, CPU-only cloud compute instances. GPU cloud compute is almost unbelievably expensive. Even Linode charges $1000 per month, or $1.50 per hour (look at the GPU plans: https://www.linode.com/pricing/#row--compute ) An AVX-512 Skylake-X cloud compute instance costs $10 per CPU-core per month at Vultr ( https://www.vultr.com/products/cloud-compute/ ), and you can do about 18 De…

How could it be competitive with GPUs in terms of price-per-unit-performance? Seems like GPUs are expensive only because you can't rent a small portion of a GPU? But shouldn't that be possible with GPU virtualisation? Or is it the case that if you virtualised a GPU up into tiny pieces, the memory-to-flops ratio would be way off what's needed for inference? Or the virtualisation overhead would be too big? Those are al…

Well, a GPU is a cluster of SIMD units with fast memory

A GPU thread variable is just like a SIMD lane, and a GPU warp variable is just like a SIMD vector register

Nvidia's SIMD instructions are called PTX and they are similar to AVX-512

An AVX-512 core is like a general purpose CPU with a 512-bit GPU core built in

So paying for a single AVX-512 core is like paying for part of a GPU, plus the general purpose compute you need to keep the GPU supplied with work

If you could divide the GPU up, you would lose most of the parallelism, keep all of the communication latency, and still need the drivers etc.

Would a hypothetical virtualized GPU be competitive with an AVX-512 core in terms of price/performance? I don't know, I haven't done the comparison

Re: Ask HN: What startup/technology is on your 'to watch' list?

#189

Post by nlstitch »

Was honestly looking for something revolutionary but did not find any here; Just some service/app to replace another service.

Where is my electric driverless car? Space travel or actual tech that prevents polarization and prevent narrow targeting of people that feel drawn to conspiracy theory rabbit holes? We should be aiming much higher.

Re: Ask HN: What startup/technology is on your 'to watch' list?

#190

Post by kgin »

ibeckermayer wrote: post #146
steve_taylor wrote: post #133

The market for self diagnostic medical technology is going to boom at some point. It will be a long time before we have an AutoDoc, but until then, diagnostic equipment and diagnosis automation are something to keep an eye on and possibly invest in. Very few doctors will see it coming, because they don't want to believe that they can be partially automated.

Except this is already heavily regulated and/or the AMA will lobby to regulate it, making it impossible/expensive in many cases, slowing innovation, and keeping doctor’s salaries and health care costs high

If so, the United States will be left as island with tech adopted by developed countries with socialized systems and developing countries with fewer regulations.
Post reply on HN