John_Rich

Thank you for the help, I was able to get it up and running.

It was only on Win7, but I was able to get Pop to work.

Thank you, I’ll try it tonight.

So, I’ve been needing a new computer for a while and decided to try to fix my old laptop. It’s an HP from 2012 or 2013, and was fairly well spec’d for the time (it could run a minecraft server and minecraft client at the same time with 3 other users and minimal “quality reductions”) i7 and 16GB RAM if I remember correctly. So, the problem, I’m pretty sure about, is a HDD crash. The drive will click when you try to access the data. It’s been dead for about 4 years now and it’s just been waiting to be repaired or scrapped (I have a problem, I know). So, I got an SSD with 480GB, and I’m thinking of running Pop_OS for the operating system, since I’ve been wanting to try it and I don’t want to buy a windows license. I can replace the hard drive, no problem, but I can’t seem to get the Pop_OS to be recognized. I’ve probably done something obviously wrong, but at this point I’m lost. My current “toolset” is a working laptop (wife’s), flash drive, and a “portable hard drive converter” that allows me to plug a hard drive in as a flash drive. I have the old drive and I plan to do data recovery on it later on, but I don’t think that will be helpful for this. If I’m not able to get this working, I’ll probably have to scrape together the cash to go buy a new one. Any thoughts or suggestions would be appreciated.

So first of all, mods if this is the wrong location for this, please direct me to the correct location. Secondly, this is currently completely theoretical, and I was hoping for some feed back on possibilities. I've been considering getting a solar power system installed in my house. It would be grid-tied with a battery back-up. For an installed system I was quoted a seemingly ridiculous amount, and I've therefore been looking into going DIY on it. In so doing, I've been looking at the lowered efficiencies at each step in the conversion of the power, like if you get 100Wh out of a panel, and then lose 5% before it hits the batteries, and then 10% in the batteries, and then 10% in the conversion to AC, and then if you are using the power in a computer(or other computing device) it gets converted to DC again at like 15% loss, you would end up with only actually using around 65Wh. Now those are just general numbers, but they demonstrate the point. So, the thought was "Could you run DC electricity to a wall outlet and use it to charge a phone, or run a PC, with maybe a voltage step-down only. Furthermore, would doing so reduce the lost power enough to make this worth it?" Obviously you would use a different connector than the standard, if only so you don't accidentally plug something in wrong. I figured that a USB outlet for charging a phone would be the easiest, but I'm really wondering about the possibility of replacing a PC power supply, with something that could change the amount of volts, but doesn't have to convert from AC to DC. Thoughts? Ideas? Suggestions? People who have done something similar? I couldn't find anything out there.

So, it's like a mini computer that is specialized to a single task and not a general all purpose thing, I would probably just say "mini computer" for short and expect people to understand that it's an approximation, not an exact statement. @GoldenLag You have a good point there, but honestly looking at those cards I was thinking more of a "how could it be done" kind of thing. I still think that it's possible, and could even give a gain to enthusiasts, however I do see why it doesn't happen. @DildorTheDecent Thanks!

Ahh, that sounds more like something that would cause this. I wonder if an indie company could do this?

Of course they do, but that's the same issue as when a cooler is removed and a water cooler is added. As far as the soldering of memory being required for a good connection, desktop CPUs aren't soldered in, and that seems like a much more serious juncture. I'm sure that some kind of "proper pressure" mounting system, like the AMD TR CPUs could be made to ensure that the VRAM was properly seated. I can't imagine that the cards are not tested before you add a cooler, it would just be far too wasteful to either trash a perfectly good cooler, or recover it. Therefore, you could just grab a stack of 100 cards without coolers between these steps. I think that you may have the real answer here, perhaps it is just too niche.

So, I was hoping that someone with more insight could answer this. Why are Graphics cards only sold as full units? I have 2 reasons for this question: 1. If the end user is going to remove the stock cooler and add a watercooler on, then you could grab cards out of your normal production line before they get a cooler added and sell them for basically the same price. You would be decreasing your costs (don't have to buy a cooler for it) and therefore increase your profit, and you'd end up making the water cooling easier on the enthusiast, by supplying a naked pcb. 2. Because the Graphics card is basically a mini-computer in its own right. Why can't we get a board that can receive a GPU and video RAM like a MB receives RAM and a CPU? There seems to be an under served market for Graphics card customizations. For example, maybe your GPU has plenty of CUs but you need more VRAM. Right now you have to do a whole new card, which is more expensive than a couple sticks of VRAM. It just seems like this is something that should exist, so why doesn't it?

Could you explain this? I'm not really understanding the concern.

Thank you! So the dual socket would be less complex and in very specific use cases, such as a monster program needing 128 cores, it could be slightly faster based on your network speed. Of course if you go through the proper set-ups and have a fast enough network connection, or are running a bunch of smaller programs, you won't get any speed boost with dual socket. It's also important to know if your program can run distributed or not, so it could or could not use a second board's resources. Did I summarize that correctly? Thank you!

So, this is just theoretical for me, but I couldn't find a satisfying answer. I'm probably needing someone like a Sysadmin to answer this, but I'd like to hear from anyone that has thoughts. So, if money, space, electricity, and heat are not an issue, and you are only concerned about raw speed would a Dual socket MB be faster than 2 motherboards? So for Example: Set-up #1: 1 X {Dual socket board, 2 Epyc 7002s, 4 TB RAM} Set-up #2: 2 X {single socket board, 1 Epyc 7002, 2 TB RAM} Anything else would follow the same scheme, which is faster? I was thinking about this because I'm wondering about MB design. Obviously there are differences in practice, not the least being the ability to hot swap a single CPU rack instead of a more expensive double CPU, but I'm only interested in performance on this. Thank you and sorry if I've completely screwed something up here.

Another idea that I haven't been able to find on the web, fluid cooled de-lidded CPU. Such as "Remove IHS and thermal paste", strap on a "water block" without a contact plate, but with a gasket, use a NON-CONDUCTIVE fluid in the loop. So that I'm clear, the non-conductive fluid would make direct contact with the CPU die. No thermal paste, no thermal interfaces, nothing. Just naked die and moving fluid. Maybe use mineral oil? Anyways, it could be a cool episode for "bad cooling ideas"!! I just can't find any info on this anywhere, but theoretically it should work?

It is odd to not accept any form of payment, especially when you could probably have a simple script written that immediately, or at a predetermined point, sold the crypto for CND and deposited it into an account. You don't even need to vocally support what Brave is doing, just say "Hey, we don't have any recommendations on crypto, but we are accepting Brave tips, which will be immediately turned into cash." Of course there could be another tax/reg/business reason to not do so. Like, a requirement in Canada that you submit all records of crypto transactions, which would be too much of a hassle for $500.

How about the "Tim Allen of workstations". I'm thinking 2 (or 4 if it's out by then) socket MB for Epyc or TR, with 2 TB of RAM per CPU, with comparable graphics. Obviously overclock everything and water cool as well, maybe with a de-lidding. It would be great if it could be the rumored 64 core Threadripper that may be out next year. Basically the kind of "low-end" supercomputer that Tim Allen would introduce as "4 socket, 256 cores, 512 threads, 8 TB of RAM, this baby can render every system in the Galaxy and laugh at you while doing it! *Grunt* *Grunt* *Grunt*"