Skip Navigation

InitialsDiceBearhttps://github.com/dicebear/dicebearhttps://creativecommons.org/publicdomain/zero/1.0/„Initials” (https://github.com/dicebear/dicebear) by „DiceBear”, licensed under „CC0 1.0” (https://creativecommons.org/publicdomain/zero/1.0/)B
Posts
0
Comments
459
Joined
3 yr. ago

  • User replaceable RAM is slow, which means you can’t integrate the CPU and GPU in one package. This means a GPU with it’s own RAM, which has huge disadvantages.

    Even a 4090 only has 24GB and slow transfers to/from VRAM. The GPU can only operate on data in VRAM, so anything you need it to work on you need to copy over the relatively slow PCIe bus to the GPU. Then once it’s done you need to copy the results back over the PCIe bus to system RAM for the CPU to be able to access it. This considerably slows down GPGPU tasks.

  • "unified memory" is an Apple marketing term for what everyone's been doing for well over a decade.

    Wrong. Unified memory (UMA) is not an Apple marketing term, it’s a description of a computer architecture that has been in use since at least the 1970’s. For example, game consoles have always used UMA.

    Every single integrated GPU in existence shares memory between the CPU and GPU; that's how they work.

    Again, wrong.

    While iGPUs have existed for PCs for a long time, they did not use a unified memory architecture. What they did was reserve a portion of the system RAM for the GPU. For example on a PC with 512MB RAM and an iGPU, 64MB may have been reserved for the GPU. The CPU then had access to 512-64 = 448MB. While they shared the same physical memory chips, they both had a separate address space. If you wanted to make a texture available to the GPU, it still had to be copied to the special reserved RAM space for the GPU and the CPU could not access that directly.

    With unified memory, both CPU and GPU share the same address space. Both can access the entire memory. No RAM is reserved purely for the GPU. If you want to make something available to the GPU, nothing needs to be copied, you just need to point to where it is in RAM. Likewise, anything done by the GPU is immediately accessible by the CPU.

    Since there is one memory pool for both, you can use RAM more efficiently. If you have a discrete GPU with 16GB VRAM, and your app only needs 8GB VRAM, that other memory just sits there being useless. Alternatively, if your app needs 24GB VRAM, you can’t run it because your GPU only has 16B, even if you have lots of system RAM available.

    With UMA you can use all the RAM you have for whatever you need it for. On an M2 Ultra with 192GB RAM you can use almost all of that for the GPU (minus a little bit that’s used for the OS and any running apps). Even on a tricked out PC with a 4090 you can’t run anything that needs more than 24GB VRAM. Want to run something where the GPU needs 180MB of memory? No problem on an M1 Ultra.

    It has nothing to do with soldering the RAM.

    It has everything to do with soldering the RAM. One of the reason iGPUs sucked, other than not using UMA, is that GPUs performance is almost limited by memory bandwidth. Compared to VRAM, standard system RAM has much, much less bandwidth causing iGPUs to be slow.

    A high-bandwidth memory bus, like a GPU needs, has a lot of connections and runs at high speeds. The only way to do this reliably is to physically place the RAM very close to the actual GPU. Why do you think GPUs do not have user-upgradable RAM?

    Soldering the RAM makes it possible to integrate a CPU and an non-sucking GPU. Go look at the inside of a PS5 or XSX and you’ll see the same thing: an APU with the RAM chips soldered to the board very close to it.

    This again has little to do with being socketed though: LPCAMM supports up to 9.6GT/s, considerably faster than what ships with the latest macs.

    LPCAMM is a very recent innovation. Engineering samples weren’t available until late last year and the first products will only hit the market later this year. Maybe this will allow for Macs with user-upgradable RAM in the future.

    The only way discrete GPUs can possibly be outcompeted is if DDR starts competing with GDDR and/or HBM in terms of bandwidth

    What use is high bandwidth memory if it’s a discrete memory pool with only a super slow PCIe bus to access it?

    Discrete VRAM is only really useful for gaming, where you can upload all the assets to VRAM in advance and data practically only flows from CPU to GPU and very little in the opposite direction. Games don’t matter to the majority of users. GPGPU is much more interesting to the general public.

  • The best I've found is benchmarks of Apple silicon vs Intel+dGPU, but that's an apples to oranges comparison.

    The thing with benchmarks is that they only show you the performance of the type of workload the benchmark is trying to emulate. That’s not very useful in this case. Current PC software is not build with this kind of architecture in mind so it was never designed to take advantage of it. In fact, it’s the exact opposite: since transferring data to/from VRAM is a huge bottleneck, software will be designed to avoid it as much as possible.

    For example: a GPU is extremely good at performing an identical operation on lots of data in parallel. The GPU can perform such an operation much, much faster than the CPU. However, copying the data to VRAM and back may add so much additional time that it still takes less time to run it on the CPU, a developer may then choose to run it on the CPU instead even if the GPU was specifically designed to handle that kind of work. On a system with UMA you would absolutely run this on the GPU.

    The same thing goes for something like AI accelerators. What PC software exists that takes advantage of such a thing?

    A good example of what happens if you design software around this kind of architecture can be found here. This is a post by a developer who worked on Affinity Photo. When they designed this software they anticipated that hardware would move towards a unified memory architecture and designed their software based on that assumption.

    When they finally got their hands on UMA hardware in the form of an M1 Max that laptop chip beat the crap out of a $6000 W6900X.

    We’re starting to see software taking advantage of these things on macOS, but the PC world still has some catching up to do. The hardware isn’t there yet, and the software always lags behind the hardware.

    I've heard about potential benefits, but without something tangible, I'm going to have to assume it's not the main driver here. If the difference is significant, we'd see more servers and workstations running soldered RAM, but AFAIK that's just not a thing.

    It’s coming, but Apple is ahead of the game by several years. The problem is that in the PC world no one has a good answer to this yet.

    Nvidia makes big, hot, power hungry discrete GPUs. They don’t have an x86 core and Windows on ARM is a joke at this point. I expect them to focus on the server-side with custom high-end AI processors and slowly move out of the desktop space.

    AMD has the best papers for desktop. They have a decent x86 core and GPU, they already make APUs. Intel is trying to get into the GPU game but has some catching up to do.

    Apple has been quietly working towards this for years. They have their UMA architecture in place, they are starting to put some serious effort into GPU performance and rumor has it that with M4 they will make some big steps in AI acceleration as well. The PC world is held back by a lot of legacy hard and software, but there will be a point where they will have to catch up or be left in the dust.

  • So that leaves performance, which I honestly haven’t found good numbers for. If you have this, I’m very interested, but since RAM speed is rarely the bottleneck in a computer (unless you have specific workloads), I’m going to assume it to be a marginal improvement.

    This is where you're mistaken. There is one thing that integrated RAM enables that makes a huge difference for performance: unified memory. GPUs code is almost always bandwidth limited, which why on a graphics card the RAM is soldered on and physically close to the GPU itself, because that is needed for the high bandwidth requirements of a GPU.

    By having everything in one package, CPU and GPU can share the same memory, which means that you eliminate any overhead of copying data to/from VRAM for GPGPU tasks. But there's more than that, unified memory doesn't just apply to the CPU and GPU, but also other accelerators that are part of the SoC. What is becoming increasingly important is AI acceleration. UMA means the neural engine can access the same memory as the CPU and GPU, and also with zero overhead.

    This is why user-replaceable RAM and discrete GPUs are going to die out. The overhead and latency of copying all that data back and forth over the relatively slow PCIe bus is just not worth it.

  • That's why I don't buy laptops with soldered RAM.

    In my opinion disadvantages of user-replaceable RAM far outweigh the advantages. The same goes for discrete GPUs. Apple moved away from this and I expect PC manufacturers to follow Apple’a move in the next decade or so, as they always do.

  • Because it’s a 9 volt battery.

    Batteries are cylindrical because it’s the most efficient shape to make them in. There is a central electrode with the electrolyte around it. By making it cylindrical it’s distributed evenly. Imagine having it square, then in the corners the layers would be thicker than on the sides.

    So that explains why the cells and normal AA batteries are cylinders. So why not have a 9 volt cylinder? That’s because the chemistry used for alkaline batteries produces 1.5 volts. A single cell, regardless of size, only produces 1.5V. So how do you get 9V out of a 1.5V battery? By putting 6 of them in series. 6 x 1.5V = 9V.

  • You have one mail delivery service, presumably random-online-shop.com isn’t providing you mail service right? That’s the same thing here.

    No, it's not. The power company is analog to Amazon or another webshop. The network company is analog to the postal service.

    Your “provider” doesn’t produce the energy.

    They actually do produce the energy, the network company doesn't. The power company owns the power plants that actually generate the electricity, or they sub-contract to a company that does (this is a simplification, in reality it's more dynamic than that).

    They, like in the packages example you used, are simply handing off the “packages” to your power provider (the mail service) who then administers and provides you with your power (delivers your mail).

    You're playing semantic games here.

    The network company only transports the energy, it's not theirs. Just like the postal service doesn't own the contents of the packages they deliver. It's briefly in their possession, but it's not their property. And just like with the postal service they are only paid for transport. If I order a €10 bag of cat litter on Amazon, the postal company gets paid €5,95 to deliver it to my door. If I order a €2000 laptop, they also get paid €5,95. If I use 1 watt-hour of energy on a particular day , Enexis gets €1,16 for delivery it. If I use 1000 watt-hours on that day, they also get €1,16.

    Let me put it this way: let's say you got a new pair of sunglasses. You meet a friend and he says, "cool shades, where did you get them?". What will you answer?

    • "I got them from the postal service"
    • "I got them on Amazon"

    And again, if all the packages are the same, there’s no difference and it doesn’t matter, why would you sign a contract with any of these companies specifically? Why doesn’t the entire NL just buy from the cheapest company if it’s all the same to you as an end-user of the network?

    Because while the actual power is the same, the contracts aren't. There is no "cheapest company", there is maybe a cheapest company for my particular situation. There are many differences between power companies and the products they offer.

    • I may want to contract with a company the only produces renewable energy (in fact, I do).
    • I may want to have a contract with a guaranteed fixed price for the next, 1, 3, or 5 years. Prices differ for these between companies due to their prediction of how their costs (to either produce or buy) will change in that period. They also change over time. So if my fixed-price contract is up today, company A may be the cheapest option to renew, but if I had to renew 4 months from now, it could be company B.
    • I may want a contract with variable pricing, where the price is adjusted based on the price of energy on the European market every 3 months.
    • I may want a completely dynamic contract, where the prices vary on an hour-by-hour basis. Even within this there are differences. For examples: some companies will just let you pay the actual market price plus a small markup. This may be risky for the consumer is there is a suddenly a huge spike in price for a few hours (which has happened in the past). Some offer dynamic pricing based different rules, e.g. an average over the last day or so. Which takes away some of the risk for spikes but also means you don't get a lot of savings when the price is suddenly super low or even negative (which has also happened in the past).
    • They provide different levels of service. For example: they offer discounts on certain power-saving products, from LED bulbs to heat exchangers.
    • If you have solar panels, you can sell the excess power they generate back to the power company. The amount they pay also varies per power company. If you barely use any power at night and you over-produce a lot during the day, you may want to choose a company not based on how much they charge for power, but how much they pay.
    • Then there's companies that offer additional stuff. For example: there was at least one that installed solar panels on your roof, and then you'd use the money you'd save in energy usage to pay for the panels in X years.
    • Other services that are offered by power companies are things like home inspection to see if there are areas where you cold save power, e.g. by improving insulation.

    So it entirely depends on your preferences as a consumer. Do you want a guaranteed price or do you want to take a little bit more risk? How much risk? What kind of service do you want? Just the bare minimum or do you want more?

  • No, but they do read their bank account statement before passing to see if the bribe campaign donation was paid in time.

  • the key difference is that most of the delivery infrastructure is fully owned and operated by the same companies that do the production

    That used to be the case here as well, but these companies got split up into a network company and a power company when the government decided to open up the energy market for competition.

    It sounds like the Netherlands does in fact have two power providers then in the form of Enexis and TenneT, at least as I understand the term "provider" as we use it here. Or do you also maintain contracts/payments with individual power production firms that you haven't already described? What is the benefit of maintaining a contract with any of these companies if they already sold the power to Enexis/TenneT and they don't do any other administration services?

    You misunderstand. Power companies do not sell their power to TenneT or Enexis. It’s the other way around: power companies pay TenneT and the local network owner (Enexis is just one of them) for transporting the power they generate to their customers. These are the fixed network costs that get passed on to the consumer.

    TenneT, Enexis and the other network companies own the power distribution infrastructure, and they basically charge for access to their network.

    Saying there is only one power company because Enexis delivers the power to my house is like saying there is only one online shop because the postal service delivers all my online orders to my door.

    Just like I can order something from Amazon, or bol.com, or CoolBlue, I can also buy power from any of these 52 power companies. It all gets transported by TenneT and Enexis just like all my packages are transported by PostNL. And just like when I order something online, I don’t have a contract with the delivery service: the delivery service is contracted by the seller. Just like the PostNL shipping costs show up on my bill, so do the energy transport costs. I still only have a contract with the energy company.

    The energy company either generates the power themselves (analog to Amazon shipping from their own warehouse) or they can pay someone else to generate the power they sold me (analog to Amazon drop-shipping a purchase). The company I currently buy my power from (Essent) has several power plants, but they may buy additional power on the open market, or sell it if they have spare capacity. This varies hour by hour.

    Since all electricity is the same, there is no need to transport the exact energy produced by my energy company to my house. Or to use the packages analogy: imagine every online shop only sold 1, identical product. Online store A sold 3 products to customer 1, 2 and 3, and online store B sold 2 products to customer 4 and 5. Since all products are exactly the same, they don’t bother putting address labels on them, company A just sends 3 unmarked packages to the postal service warehouse, and company B sends 2 unmarked packages. Since it’s all the same, the postal service just tosses all the incoming packages onto one big pile. They then deliver a random package from that pile to customers 1 through 5, it doesn’t matter which as they’re all the same anyway. Why keep track when there is no point? The only thing that matters is that if company A tells the postal service to deliver 3 packages to their customers, they should also send 3 packages to the postal services’ warehouse. No need to keep track of individual packages, but number of packages in and out must match for each online store.

    We regularly experience outages affecting people into the hundreds of thousands.

    Power outages are extremely rare here and usually only affect a small group of houses. With the exception of the high-current linking network, all our power lines are underground so they aren’t affected by things like falling trees and the like.

  • Or you suddenly find health kits and ammo everywhere.

  • How does this work? I looked and according to the Internet TenneT is the sole operator of the national electrical grid in the Netherlands. Do you have 52 power lines running to your home?

    TenneT runs the what is called the ‘koppelnet’ (linking network). This is a high-voltage network that links the power stations together and also links to the europe-wide electrical high voltage network. Locally, the distribution networks form the link between the high-voltage net and your home. It first goes to a medium-voltage network and finally a low-voltage network. In my case the distribution network is owned by Enexis.

    Neither Enexis or TenneT produce power. TenneT is owned by the government, Enexis is a private company but these are heavily regulated (e.g. there are caps on how much profit they are allowed to make).

    As a consumer, you pay a fixed amount per day for use of the distribution network. You pay this through your normal energy bill. This is specified separately on the bill. In my case this means I pay €426,95 a year to Enexis. This fee is purely based on the size of my connection to the grid, in my case 3 phase 25 amps. It doesn’t change if I use more or less power.

    As for where my electricity comes from? I don’t know, because it changes constantly and with how electricity works, I’m not even sure if this question makes sense. All power producers together keep the grid supplied with enough power. Power stations across the country are linked together and how the generated power is distributed is monitored and adjusted constantly. This also includes capacity from other European countries, as those grids are linked too.

    Lots of people have solar panels nowadays and if they produce more than they need they can supply that back to the grid as well and get paid for it (it basically means their electricity meter runs backwards).

    What do these companies do if they don't actually produce the electricity or operate the grid?

    A bunch of them do produce the electricity, but as I said the power plants are connected to the nation-wide linking network. There is no direct relation between any of these power plants and my home. They can also pay someone else to produce it for them. This could be someone in a completely different country. A bunch is produced locally through solar panels and sold to the power company by consumers.

    Some smaller companies may not produce anything themselves and just buy it somewhere, but where als can vary day by day and hour by hour. They will buy it wherever they can get the best deal at that time.

    All that matters is that they make sure they provide enough energy to the national grid to cover the use of their customers. How it’s actually distributed is completely separate.

    Note that due to how the energy market works, electricity prices change hour by hour. In my case I have a contract with a fixed price for a certain amount of time (1 year, but you can get longer contracts), but it’s also possible as a consumer to get a completely flexible contract. That means the price you pay for electricity changes every hour, based on the actual market price (plus a markup for the energy company of course). Especially during the day when lots of solar power is available, the price can drop a lot. There have even been cases where the price was negative for a short while (i.e. you got paid to use electricity) due to overproduction. If you have a flex contract, your energy company usually has an app that will tell you the current price and the expected price for the next couple of hours.

    I have precisely one power company that services about 5 million residents here where I live. No other options.

    So if their power plant goes down 5 million people are without power? That sounds extremely fragile.

  • I'm an IT person professionally, and I use Fedora as my daily driver.

    Ah, Fedora, that brings back memories. We used to call it RootHat back in the day when it was still RedHat. It was what all the first-time Linux users used before they graduated to Debian or Slackware. They would use root as they day to day account, hence the name.

    Havent used it in forever. Is it still as big a pile of shit as it was in the 90’s ?

  • but you probably can put your tools in the bed

    But that’s still impractical, you can only put your tools in there temporarily, but you can’t leave them unattended in an open truck bed or they would get stolen. Over here carpenters, masons, electricians and people in similar occupations usually use vans. Often they have shelves and other storage solutions in their van for their tools. Here are some examples (text is in Dutch but the photos are self explanatory).

    You can still use a trailer with a van, and they often have roof-racks for things like long pieces of lumber, a ladder, etc. In comparison a truck just seems way less practical than a van.

  • You’re choosing from vendors to deal with the regional vendor. You’re just paying someone to pay the company in your area.

    That's not how it works.

    The price you pay has 2 main components: a fixed component for network costs. This part is always the same and only depends on the type of hookup you have (e.g. an industrial power connection would cost more than a residential one). It goes to the owner of the physical power network (this depends on where you live) who uses it for things like maintenance (and of course a little profit). The network operator does not provide power themselves.

    The variable component is based on your usage and goes to the power company. The power companies provide the actual power to the national grid. While it may be true that the power coming out of my outlets wasn't produced by the company I'm contracted with, that doesn't really matter. Electricity is electricity. What matters is that each power company has to provide the amount of energy used by their customers to the national grid. Say company A has 100 customers that on average used 1 kilowatt-hour each, and company B has 50 customers that used the same amount on average. Then company A has to provide 100 kilowatt-hours to the national grid and company B has to provide 50 kilowatt-hours.

    How they provide that power is up to them, and usually varies. The larger energy companies have their own power plants, wind and solar farms. Smaller companies may buy energy in bulk from the larger ones and try to sell it at a profit to consumers. Energy companies may also buy or sell to/from other countries, depending on capacity and demand.

    Point is that they don't just forward you the bill from a local company, they actually have to provide the power and outside of a few small 'virtual' energy companies they do produce that power.

  • And yet, I can choose from dozens of different energy companies for electricity and gas.

  • What exactly is the great advantage of eSIMs if you have a dual SIM phone?!

    eSim means you don’t have to go to a store to get a physical SIM. You can use a ‘SIM store’ app to get an eSIM for wherever you are.

    Another minor advantage is that you don’t need a SIM PIN as the SIM is a physical part of the phone. So you only need to enter one code when you restart your phone.

  • Yeah, I usually see farmers use a 4WD with a trailer, or just their tractors. Trucks aren’t really a thing here, just the occasional douchebag who has imported one from the US.

    For uses like construction and other blue collar work people use vans. Lots of storage space, and it’s enclosed so protected against the elements and from theft.

  • What does one need a truck for anyway? They seen highly unpractical as working vehicles.