Der8auer’s video is worth a watch, he got one of the Redditor’s card:

https://youtu.be/Ndmoi1s0ZaY

Yes, so R&D and finalizing the model weight is done on NVIDIA GPUs (I guess you need an excessive amount of VRAM).

Inference is probably gonna be offloaded to consumers in the end where the NPU is taking care of the inference cost (See Apple, Qualcomm etc)

Not the best on AI/LLM terms, but I assume that training the models was done on Nvidia, while inference (using the model/getting the data from the model) is done on Huawei chips

To add: Training the model is a huge single-cost expense, while inference is a continuous expense.

Sounds like you might just be max’ing out the capacity of the coax cable as well (depending on length/signal integrity). E.g. ITGoat (not sure how trustworthy this webpage is, just an example) lists 1 Gbps as the maximum for coax while you would typically expecting less than that, again depending on your situation (cable length, material, etc)

What’s your situation into the wall? Depending on country/ISP/regulations they might give you up to 1000 Mbps under the assumption that it’s a single line going to a single user, however quite often that line is shared with potentially a lot of different customers.

Some countries allows you to buy packages where you have a standalone line going to your wall, however at an additional cost

No bios update, but you most likely received both microcode updates (which is what will fix/mitigate the Intel issue, the bios is only to ensure everybody gets the microcode update) and firmware updates (from linux-firmware)

Of course non-mainlined (i.e. not in the linux kernel) firmware is a bit more iffy, luckily it’s getting slowly better with OEMs using fwupd for those scenarios

Breaking Linux every week or every other week? That’s almost impressive!