[astrange]

> - 5G (still no SV grown solution there)

Is this a real thing that anyone cares about? There was a lot of marketing from Samsung and Supermicro for a few months about how everyone needed "5G AI cloud infrastructure", but it seemed like an especially bad strain of bullshit. I certainly haven't figured out what it was supposed to mean yet.

Selling infrastructure to ISPs is not an amazingly high margin business.

[onethought]

Yeah, 5G is a real thing. It’s really fast and awesome!

It involves both improvements in hardware and software. Huawei are the leaders in it. (Devastating to the narrative that China just copies everyone)

[astrange]

5G FR2 is certainly fast but we already had "fast" (5Ghz Wi-Fi or wired Internet), and it's pretty harsh on mobile batteries. 5G FR1 is somewhat better than LTE but not really enough to enable new uses.

The real issue preventing new consumer uses is that wireless plans are expensive and have data limits. Driving that down presumably involves investing in the wireless ISP backend, which doesn't have much to do with the radio technology.

[inshadows]

Define awesome. How will it improve my life? Please don't mention high speed. Or if you do, mention applications where it makes difference.

[throwaway-8c93]

I do share your skepticism overall, 5G is being pushed more aggressively than its value would warrant. But there are parts of 5G suite that are indeed valuable and go beyond the mere "moar speed" mantra.

Network slicing is not consumer facing, but is a big deal in business-to-business connectivity, likely enabling business models that just aren't possible today. It's a bit like renting virtual machines in the cloud, but instead you're renting connectivity, tailored to your SLA needs.

Higher bandwidth, lower latency and lower power consumption are not exciting in isolation, but improving all of them at the same time does bring notably better user experience (see the raving reviews of M1 Mac for the same phenomenon).