• 2 Posts
  • 932 Comments
Joined 3 years ago
cake
Cake day: August 16th, 2023

help-circle



  • It certainly seems like your GPU is causing a crash, but it’s just as likely (maybe even moreso) that the PSU is behind that.

    You can look into things like Furmark, MATS and MODS, etc to drill down further. Northwest Repair has plenty of examples on his YT channel.

    Is this a new issue, or has it always been the case with this hardware combo?

    If it were me, I’d just get a new (and good) PSU from somewhere with a good return policy. If that doesn’t fix it, return the PSU for a GPU. Other people will link to the calculators, and be sure to get a good model from a good brand. Don’t cheap out or ignore the subtle details. A bad PSU will pay you back with all sorts of weird, intermittent stability issues.



  • The answer really depends on your needs and use cases, as well as your risk profile.

    The simplest option is to schedule regular backups to a USB (preferably rotating drives, following the 3-2-1 rule). If “shit happens”, you would restore that backup using the same software. You may have difficulties restoring to different hardware. Acronis and Macrium are simple solutions here, or Clonezilla if you prefer free or offline backups.

    If you want a way to boot to a diagnostic tool, and recover from there, there are countless options. From a live Linux distro, UBCD, WinRE, Media Creation Tool, the options abound. You may or may not be able to recover any data, or your existing system. It really depends on your definition of “shit happens”, and what you need it to do.

    I suspect that your real goal is to move your existing Windows install (and all associated data) to a USB drive, and boot from that. Windows to Go was exactly that, but was discontinued. There are some hobby projects to do the same. This is (usually) a really bad idea. USB drives are much slower and less reliable, and create extra points of failure. It also keeps almost all of the other risks of the internal drive. You’re better off using option #1 above.



  • Not OP, but I think they’re trying to make a (strained) point about companies replacing highly paid professionals with amateurs using technology. There’s also a side note of said technology being just as good, if not better than a skilled and experienced craftsman.

    Of course, I could be wrong. The analogy has many flaws, and I’m making a lot of assumptions.


  • Dan Sullivan isn’t exactly a unique name. I bet there are thousands of people in the US with that name, and at least dozens in Alaska alone. Why should one guy get to own that name for elections, and disqualify any challengers who share the same name? What about even more common names like John Smith?

    I get that it creates a unique challenge to design the ballot so that it’s clear. I also don’t see that as a difficult task, given that they have different middle initials and one will be listed as the incumbent. Their campaigns will probably emphasize the initials, with something like “The S is for Serious” (except better).






  • Focus on the business needs. What do you need? Why do you/they need it? What will they have that they don’t have now? This part is very important. If you can’t come up with a compelling business case, then there’s no reason to move forward on it.

    How much will it cost? Are there any cheaper options? Some places would rather hire a $100k/year sysadmin rather than buy a $10k/year license. Other places will get excited when you talk about reducing support staff. Keep that in mind. Also, some places are more concerned about how replaceable someone is, so using standard products are more valuable than price or capability.

    When you’re evaluating options to recommend, try to get information about the entire landscape. There’s a very good chance that they’re already paying for Microsoft E5, which means you already have a license for Intune that you aren’t using. If not, you can pitch it as a total package with other benefits.

    ETA: Security is a really big deal at most companies. Internal and external threats, data loss, exfiltration, legal compliance requirements, etc. But smaller places often won’t care until they’ve been burned.



  • Huge is subjective, but there are many solar farms today running on just a few acres. Some quick googling says that community-based and commercial farms typically run on 10-40 acres, generating between 50-200 megawatts. If you consider 40 acres to be a huge farm, then sure. But it doesn’t make much sense to run these in a decent populated area anyway. This is no different than coal, gas, or other power plants today, albeit for different reasons.

    As for huge batteries, yes, we will need them once solar power capacity reaches a point where generation in the day exceeds the demand. As it stands, many places still have off-peak pricing at night. Solar would need to counterbalance that completely before we need huge batteries for solar.