deleted by creator

deleted by creator

deleted by creator

deleted by creator

I’m sure the teacher of the high school debate team you’re on would be very proud ❤️

ooh getting aggressive now are we?

I owe nothing to you. Enjoy your time being a sad person trying to bring others down on the internet :) I hope this little outlet makes you feel better

Sure, buy an inverter and burn up 10% of your energy in the conversion if you’re lucky. That inverter will cost roughly as much as the contents of a standard fridge + freezer, by the way :)

At that point just buy a well insulated cooler and always have some ice on hand. It’ll last much longer.

Congratulations, this is the worst attempt at ridicule I’ve ever seen

Now I don’t know enough about electronics to know how wrong this is

Very, assuming the refrigerator in question typically runs on a typical power grid you’d find in the US or Europe (source: am electrical engineer)

Mainly because most compressors I’m aware of use alternating current (AC) motors, or at a minimum accept AC power. Batteries alone produce direct current (DC). The simplest way to make this work would involve an inverter (converts DC to AC). Cheap ones probably have at least a 10% conversion loss, so you’re looking at an hour or two at most.

Edit: should also mention that discharging a typical lead-acid battery until it’s all the way flat (realistically below ~11V) does irreparable damage. Might be cheaper to replace the contents of your fridge :)

deleted by creator

This is especially true in the research space (where 90% of this stuff is being built :)

Right? TPUs make more sense at scale (especially for LLMs & similar). The consumer market is more about hype and being a household name than it is about revenue.

Couldn’t agree more! Abstracting to a general economic case – those hundreds of dollars are a double digit percentage of the overall cost! Double digit % cost increase for single digit % performance doesn’t quite add up @nvidia :)

Especially with Google going with TPUs for their AI monstrosities it makes less and less sense at large scale for a consumers to pay the Nvidia tax just for CUDA compatibility. Especially with the entrance of things like SYCL that help programmers avoid vendor lock.

I like the per-application fidelity that you get from RethinkDNS. You’re definitely right though, PiHole is a great solution! Personally I run PfSense in a VM and PiHole in a container on the same hypervisor so there’s roughly no latency.

The Jetson solution on the other hand is perfect for hotels/dorm rooms/etc. Both options are excellent when used well!

Or run Android TV (Lineage OS) on a Jetson Nano with RethinkDNS to get a high quality experience you actually have control of.

Unfortunately I’d say these fall in the consumables category. Realistically the failure mode of a hiking pole is bending or cracking (not as easy as swapping a part) and the failure mode of micro spikes is the flexible part breaking or the spikes wearing down. I’m tuning in to see if anyone disagrees though :)