I have no clue what MultiPlex theaters are

So back in the earlier days of cinema, you’d go to the Cineplex to see a movie. A Cineplex would only have a single screen for viewing movies while the multiplex would have multiple screens for seeing movies on. This started with the first duplex theatre in 1915 and later the first triplex in 1966, shortly followed by theatres with 6+ screens which is around when the term “multiplex” started being used. Basically for anyone born after the 80s (therefore anyone under the age of about 40) the term is largely obsolete since most theatres have at least 4 screens and qualify as multiplexes, plus the industry has seen so much consolidation that smaller independent theatres with 1-2 screens are pretty uncommon now

Y’know what, I honestly haven’t looked at what the PCIe lane layout is like on newer chipsets. Maybe it’s gotten better since I last really paid attention like 5+ years ago. I remember in early-mid AM4 there was a lot of grumbling about how there’s only 20 PCIe 3 lanes followed by early PCIe 4 platforms that would give only 16-20 lanes with another 8 or so PCI 3 lanes. I also didn’t really pay much attention to AMD before AM4 given how far behind Intel they were. But I could be entirely out of date now that I think about it

So hardware that may still be perfectly usable but predates NVMe should be tossed out then?

The exact same thing you already have to do to upgrade the memory on such a computer, you go buy used/old stock DDR3 or cannibalize from another system. Pre-NVMe systems are DDR3 era and older. Time goes on, interfaces update and anyone looking for compatibility with their older system will need to either use an adapter or buy used/old stock. If there’s enough demand like with motherboards there can be a random Chinese brands making new hardware for old platforms using a mix of new and cannibalized parts

I did similar when preparing my wife and I for windows 10 EOL. I went back to Linux on the new drive, my wife to Windows 11. Honestly both have a similar amount of issues (mostly wake from sleep challenges on Linux, although my PC wasn’t great about waking from sleep on Windows to begin with) and most importantly my wife can still play Fortnite and I can have fun trying new stuff out and reveling at how every single game I try just works on Linux whereas 5 years ago it was more of a 50/50 chance whether or not a game would work

Does require you to have the PCIe lanes for it, BIOS support for booting to PCIe (which Intel 6th gen core CPUs were the first to support. 4th gen never did but some had m.2 slots and NVMe support for secondary drives and the 5th gen X99s had some receive BIOS updates to support but that’s its own can of worms) and both Intel and AMD have historically been pretty bad about being stingy about PCIe lane availability

Plus to run more than a single NVMe on a single slot your motherboard either needs to support PCIe bifurcation which is almost exclusively an enterprise feature or they need to have the right lane configuration available to support that x16 slot handing out 4x4 lanes (or 2x8/2x4 for dual NVMe)

Yeah my recent IT experience is similar. I redeployed monitors that had “vista-ready” badges on them during the monitor shortages of 2021-2 I’ve replaced so many of those analogue to digital adapters (usually because the computer only has 1 digital output and 2 displays to drive, or 1 HDMI and 1 DisplayPort but the displays only support HDMI and I only have VGA to HDMI adapters, etc.)

The challenge simply comes down to the fact that displays tend to last so much longer than the computers they’re connected to. Heck my wife is using my decade old 1080p monitors because they were an upgrade over the even older 720p monitors she had before which may well find themselves mated up to my kids’ new computer

I’d ask if you’re at my work, but this is an amount of organizational improvement that they haven’t yet been able to begin

For an rail network that runs 24/7 they’re going to have crews specifically to wake up should there be a problem on the busiest sections of mainline as this hoax indicated there were. That’s a significant amount of dollars burned if they close the line due to a citizen reporting heavy damage to the bridge, and just waiting until 8am on the next business day to actually look at anything.

I strongly suspect what happened was they woke up their on-call inspectors (or scrambled an inspector who worked nights, which a rail network may very well have) informed them of photos circulating showing significant structural damage to this 150 year old viaduct, so they roll up and see the exact same viaduct in the exact same shape it’s been in for their entire life and call up their boss and say “oy you wakin me up for this shiv? The bridge is bloody fine! Check your sauces mate!” (And after reporting that it was a hoax probably went and did a more thorough inspection to make sure their bases were covered)

Jesus it’s not even like it just ran rmdir from the current directory and got lost, it just flat out explicitly deleted everything from the root of the drive

Yeah as an admin I love that I can run familiar Linuxy commands in powershell but I also hate that they can’t just use/fork the real userland utilities so everything works just similarly enough to completely throw you off when you stumble across a difference

I love reading this as the backups have never succeeded

Its all creepy until you realize it was all just a chat with an LLM and not actually an agentic machine learning model or chain of models hooked into some custom APIs

LLMs famously collapse into rediculousness once a conversation goes on too long. They’re now at the point where that takes more than a couple of paragraphs of text at least

On the upside, they specified D: drive which is typically a lesser used bulk storage drive, so less activity to potentially overwrite the files marked as deleted

Hey it’s never too late to get things better under control! My parents only just started that journey and they’re old enough to be grandparents

Honestly the questions you’re posing require a level of market analysis that could fill an entire white paper and be sold for way more money than I want to think about. Its a level of market analysis I don’t want to dive into. My gut instinct from having worked in the tech industry, working with datacenters and datacenter hardware at large companies is that the AI industry will contract significantly when the bubble pops. I’m sure I could find real data to support this prediction but the level of analysis that would require and the hours of work are simply more than it’s worth for an internet comment.

You have factors including what hardware is being deployed to meet AI bubble demand, how the networking might be setup differently for AI compared to general GPU compute, who is deploying what hardware, what the baseline demand for GPU compute is if you simulate no present AI bubble, etc. etc. it’s super neat data analysis but I ain’t got the time nor appetite for that right now

That’s super interesting! I didn’t know other languages handled side thoughts better!

Oh yeah machine learning as a technology will survive, and eventually it will be implemented where it can do what it’s really good at, but right now it’s being shoved into everything to do things it isn’t good at, so you end up with a super expensive to run, energy inefficient tool that runs worse than with traditional algorithms that can be run client side or on a single much cheaper server (I’m oversimplifying the server architecture for brevity)

Think customer service chatbots on ever car dealership’s website. Traditionally these were extremely simplistic and usually just had canned responses based on keywords in the customer’s written message and would quickly cascade the customer to a real customer service rep as soon as things got out of scope. Now with LLMs companies are running those as the customer service chatbots and the LLM can do anything from agreeing to sell a new car for a dollar to providing scam or invalid contact info to referring the customer to a competitor. There’s no knowing what the AI will do because it’s non-deterministic and you don’t want that in customer service!

Right now we’re in the bubble phase where every single company is finding some way to shoehorn AI into its business model so they can brag about it. Fucking Logitech added a remappable AI button that brings up a ChatGPT interface and just spends Logitech’s money on tokens with ChatGPT. That’s pure bubble behavior. Once the bubble pops we won’t have literally every single time you open a car dealership page spending an LLM token or 5, you won’t have Amazon running AI chatbots on every product page just for asking about that product, you won’t have every website just giving away free unrestricted access to LLMs. That’s what I’m talking about.

AI demand will drop when the bubble pops, and while it will be higher than it was 8 years ago, everyone is going to be very skeptical of anything AI, just like folks are still skeptical of mortgage backed securities over 15 years later, or just like people are skeptical of commerical websites without a clear method of financing 25 years after the dotcom bubble. People remember these things and will take a while to warm up to the idea again

Is there enough demand for thousands of servers with purpose built ARM processors (which may or may not have any publicly available kernel support) driving 4-8 600w a pop Nvidia datacenter chips though? Yes some will be repurposed but there simply won’t be the demand to fill immediately. Realistically what will happen is companies operating these datacenters will liquidate the racks, probably liquidate some of the datacenters entirely and thousands of servers will hit the secondhand market for next to nothing. While some datacenter structure city empty and unmaintained until they’re either bought up to be repurposed, bought up to be refurbished and brought back into datacenter use of torn down, just like an empty Super Walmart location

Some of the datacenters will be reworked for general compute, maybe a couple will maintain some AI capacity, but given the sheer quantity of compute being stood up for the AI bubble and the sheer scale of the bubble, basically every major tech company is likely to shrink significantly when the bubble pops, since we’re talking companies that currently have market caps measured in trillions, and literally a make up full quarter of the entire value of the New York Stock Exchange, it’s going to be a bloodbath.

Remember how small the AI field was 6 years ago? It was purely the domain of academic research, fighting for scraps outside of a handful of companies big enough to invest in am AI engineer or two on the off chance they could make something useful for them. We’re probably looking at a correction back down to nearly that scale. People who have drank the coolaid will wake up one day and realize how shit the output of generative AI is compared to the average professional’s human work

Machine learning models have much different needs that crypto. Both run well on gaming GPUs and both run even better on much higher end GPUs, but ultimately machine learning models really really need fast memory because it loads the entire weights into graphics memory for processing. There’s some tools which will push it to system memory but these models are latency sensitive so crossing the CPU bus to pass 10s of gigabytes of data between the GPU and system memory is too much latency.

Machine learning also has the aspect of training vs inference, where the training portion will take a long time, will take less time with more/faster compute and you simply can’t do anything with the model while it’s training, meanwhile inference is still compute heavy it doesn’t require anywhere near as much as the training phase. So organizations will typically rent as much hardware as possible for the training phase to try to get the model running as quickly as possible so they can move on to making money as quickly as possible.

In terms of GPU availability this means they’re going to target high end GPUs, such as packing AI developer stations full of 4090s and whatever the heck Nvidia replaced the Tesla series with. Some of the new SOCs which have shared system/vram such as AMD’s and Apple’s new SOCs also fill a niche for AI developer and AI enthusiasts too since that enables large amounts of high speed video memory for relatively low cost. Realistically the biggest impact that AI is having on the Gaming GPU space is it’s changing the calculation that AMD, Nvidia and Intel are making when planning out their SKUs, so they’re likely being stingy on GPU memory specs for lower end GPUs to try to push anyone with specific AI models they’re looking to run to much more expensive GPUs

Huh I’d never looked into that! Thank you!

I mostly used zram because there’s an easy-peasy script in the Debian repos making it dead simple to setup and never looked deeper into it