Initially a lot of the AI was getting trained on lower class GPUs and none of these AI special cards/blades existed. The problem is that the problems are quite large and hence require a lot of VRAM to work on or you split it and pay enormous latency penalties going across the network. Putting it all into one giant package costs a lot more but it also performs a lot better, because AI is not an embarrassingly parallel problem that can be easily split across many GPUs without penalty. So the goal is often to reduce the number of GPUs you need to get a result quickly enough and it brings its own set of problems of power density in server racks.

I use a 5600g on b450 ITX board and 4x 8GB Seagate drives and see about 35W idle and about 40W average. It used to be 45W because I was forced to use a GPU in addition to a 3600 to boot (even though its headless, just a bad bios setup that I can’t fix) and getting a CPU with graphics dropped my idle consumption quite a bit. I suspect the extra wattage for your machine is probably the bigger motherboard and the less efficient CPU.

It is possible to get the machine part down into single digits wattage and then about 5W a drive is the floor without spinning them down, so the minimum you could likely see with a much less powerful CPU is about 30-35W.

The thing is peertube wont grow unless the people aware of it start advertising and using it as an alternative. It takes collective investment in building the audience on an alternative for it to become viable.

Make sure none of the exceptions are ticked and the Minimum number of articles to keep per feed is also 25 or below. Then its up to the cron when that runs so you might have to manually purge it and optimise the database to see what it will actually keep.

I can’t say I have ever worried about it, been running FreshRSS for years and it seems to keep its database size in check fairly well and the defaults have worked fine for me and it rarely gets above 100MB. So I know it “loosely” works in that old articles are absolutely getting purged in time but have no idea how strictly it follows these rules.

The DMZ for the ISPs router forward to the second router, then everything that hits your outside IP will be forwarded to router 2. Then on Router 2 you open the ports for your service and forward to the internal machine. That should all work fine.

Its quite complicated to setup as well, just went through the instructions and its a long way from just add to docker and run unfortunately. Would be nice to be able to just get a runner in the same or different docker and it just works easily without a lot of manual setup in Linux of directories and users and pipes etc.

I did the same move from contabo to Netcup. Contabo I had all sorts of weird bandwidth limiting problems that I couldn’t explain and which the continued to deny they were throttling. Netcup worked perfectly.

OpenWRT supports two devices for wifi 7 on filogic 880 at the moment in snapshot, the banana R4 with BE14000 wifi card add on, its a development board, and the Asus BT8. There are still plenty of issues but Wifi 7 is starting to come to OpenWRT and these MT7988 devices are going to be the major thing supported first.

The problem is the information asymmetry, there is always another person for a fraudulent company to exploit due to a dysfunctionally expensive court system. Its why we need market level regulations and public institutions that recover peoples money and fine the organisations for their breaches. This sort of thing works a lot better in the EU than in the US due to the sales laws, the ability to return within 2 weeks, default warranty on goods out to 12 months and expectations of goods to be as advertised forced onto the retailers. They work, they need more enforcement from regulatory bodies but retailers do follow them for the most part and quickly change tune when you go to take legal action when they don’t because courts know these laws inside and out.

Mine are only 25k hours or so, around 3 years. My prior set of disks had a single failure at 6 years but I replaced them all and went to bigger capacity. There is also the power saving aspect of going down to 2 drives as well, it definitely saves some power not spinning 4 extra drives all the time.

I sub to channels and use Youtubes recommendations and new for you to find additional channels etc but I don’t watch them I use Metube and a browser plugin and download the videos to a directory. I don’t get all the privacy but I also am not giving them much watch data and I can avoid the ads.

Ultimately Google is going to do this to every customer in the end, its systems for detecting problems are buggy and you don’t get to speak to a human to fix the problem so a simple automated error becomes fatal. Its going to happen at some point and its a lot easier to migrate when it all still works than afterwards when it doesn’t.

I noticed today searching that the date search no longer seems to work right. There are some terms that only appeared since 2020 and up until my recent attempts those terms produced no results on DDG when date constrained but now produce terms in articles clearly after that date. I don’t know if this is some personalisation nonsense or always pulling but results if the constraints don’t match or what but its seriously problematic and means I can’t trust the date constraints anymore.

AMD has unfortunately a long history of abandoning products before its reasonable on its graphics division. Its not really acceptable, up until earlier this year my NAS/server was running a 3600 and its only for power saving purposes I changed that as its still a very workable CPU in that role.

They are going to get sued for billions and this little stunt isn’t going to change that. Should have implemented proper software testing before you took ever corporate computer in the world, but companies like this always force their developers to rush instead of do the right thing and when it bites them expect that things will carry on as normal. I can’t see many renewals in their future.

Either Intel is extremely incompetent and genuinely has no idea what is causing this problem or it knows full well is doing all it can to drag this out and avoid recalls. This saga has been going on all year now this is getting ridiculous.

It really depends on the project. Some of them take breaking changes seriously and don’t do them and auto migrate and others will throw them out on “minor” number releases and there might be a lot of breaking changes but you only run into one that impacts you occasionally. I typically don’t want containers that are going to be a lot of work to keep up to date so I jettison projects that have unreliable releases for whatever reason and if they put out a breaking change its a good time to re evaluate whether I want that container at all and look at alternatives.

So no its not safe, but depending on the project it actually can be.

Ideally for your router you want something that runs an open source firmware (OpenWRT, DD-WRT, OPNSense, FreshTomato). Its better because you get a completely unlocked everything you need system with security patches for the hardware’s true lifetime. Every router company stops with the security updates after a few years and then at some point it becomes part of a bot net or one of this mass hack events. Its best not to play in that game and instead run some open source firmware from the outset.

The best way to start is to look at the website for openwrt.org and use their filtering to find a device that supports your needs (at least 5 LAN ethernet ports I guess and some wifi but AC sounds like it will do). The other option is a more typical 4 LAN port router which will give you a lot more options and then add a switch to that, doesn’t sound like you care too much about it being managed or >1gbps so they are also dirt cheap.

Metube is great and also has some browser plugins as well so it’s as simple as right click and send to metube on a YouTube link. Wish there was one for android too but alas not yet that I have seen.

I don’t think modern Raspberry pi’s make much sense unless you are using GPIOs or really need the low power consumption. The 3 and the 4 were OK price wise but the pi 5 is quite close to all these N100 mini computers and they are a lot more performance and expansion compared to a raspberry pi 5 and still quite low power.

Either a Topton or similar N100 based machine or a mini PC second hand is the way to go at the ~$100 mark. The mini PC will be faster and probably more expandable and cheaper but also more power consumption.