Sockets, Sparks, and Magic Smoke - Dylan Beattie - NDC London 2023

[@matthewgiallourakis7645]

New Dylan Beattie presentation, heck yeah!

[@aguluman]

Makes me cheery

[@soganox]

I am a simple man. I see Dylan Beattie, I click to watch. Never disappoints :)

[@Dominik-K]

It truly never disappoints, always a delight to see

[@HeilTec]

Dylan Beattie is always entertaining. He always entrances the audience.

[@ARiverSystem]

Fun fact, in Mary Shelley's book Frankenstein there is actually no mention of lightning at all. It's a common trope in current movies and stuff, but it wasn't in the book. It's not the only way in which it's surprising either, the idea of Frankenstein got changed quite a bit over the years in pop culture. Makes for a really interesting read.

[@JamesLewis2]

He will be with you on your wedding-night.

[@blackasthesky]

34:23 we have seen a retail CPU clocked at more than 5GHz -- Ryzen 7000 all boost above 5GHz, same for Intel 13th gen. Base clock is lower, though. Nevertheless, it clocks like that out of the factory.

[@unicorn_tamer]

Another inspiring talk! Thank you Dylan!

[@xdata3267]

I've binged two of his presentations so far and will likely look for a third.

[@creativecraft_mc]

same

[@PaulRedeemed]

Great way to start my day! Coffee and this before work

[@sundhaug92]

34:25 We already have seen retail CPUs clocked at more than 5 GHz, such as the IBM z14 (2017) and Intel Core i9-13900KS has a max boost clock out of the box of 6 GHz

[@turun_ambartanen]

He's absolutely great at talking, but there are a bazillion tiny mistakes or slight misuses of words like this in the talk that irritate me to no end.

[@Rx7man]

Yeah, the FX9590 was released in 2013 and was clocked at 4.7ghz.. I'm still running it, it's a power hungry SOB at 220w TDP

[@dieSpinnt]

The presented train of thought, the history and the graphics including the science around Moore's "Law" do not speak of short time measurements. We are talking about a steady and stable clock (including for ALL cores in a multi-core design) here. You and others hallucinated that extra condition ("boost") out of nowhere. To be clear, let me do the same kind of cheap trick and I will boast with a processor that runs at 12GHz ... no I won't tell you for how long ... that is my little dirty secret. Good idea and nice try but ... your statement is just a lie. At least you get a **facepalm** for that:)

[@L1m3r]

@@turun_ambartanen His talk is more fitting for a comedian - @ a tech conference he should've at least included a disclaimer indicating that.

[@Rx7man]

@@L1m3r then you really shouldn't watch his presentation "The web that never was"

[@hansbaeker9769]

Just starting to watch and am wondering if he will address the question we all want answered -- what did they call electric eels before the discovery of electricity?

[@ari_archer]

dylan my beloved ty for ur talks

[@AdrX003]

Energy & Dylan, big fan!

[@maxcurzi]

small correction 11:30 : AM / FM Radio operate at a much lower frequency (~1000kHz AM to 108MHz for FM) than WiFi (2.4GHz)

[@wich1]

The graphic was ordered by wave-length, not by frequency, he could’ve been more clear on that

[@alexanderkalashnikov2067]

@@wich1 It does not matter. WL and frequency are the same thing but in different domains. That was a real mistake.

[@wich1]

@@alexanderkalashnikov2067 there is no mistake, at the start he mentions wave length and keeps talking of “higher wavelengths”. “Longer wavelengths” would be less confusing. It also doesn’t help that he calls out the frequency of WiFi rather than its wavelength, which is understandable as everyone knows 2.4GHz, but nobody knows its wavelength, but it adds to the confusion, but still no mistake in sight.

[@alexanderkalashnikov2067]

@@wich1 literally. Radio(receiver showed on slide after microwave Owen)works on 1khz-500mhz. Microwave and wifi on 2400mhz. Even thing regarding dinosaurs ended in oil. Just listen to him. Jeees.

[@wich1]

@@alexanderkalashnikov2067 yes, that is 100% correct radio has a longer wavelength than microwave and therefore goes after the microwave, no mistake. Again wavelength and frequency are related, but not the same. Higher frequency means shorter wavelength.

[@blu3tan]

he never disappoints indeed, legend

[@miszcz310]

Great talk. Only one small thing is that LEDs are analog. However, this cheapest dimmers are not. The reason being, the price only. You can make complete continuous dimmer, but you need DC power source with stabilized current. And it's kind of complicated to get it from high voltage ac.

[@mattymerr701]

He is correct that non-dimmer compatible LED globes are digital _as a system_. Either there is enough power to glow, or it browns out. It's more to do with the LED globe power supplies. They aren't built to vary their output based on the input, they are designed simply to regulate to a certain voltage. Dimmer compatable LED globes actually do that interpretation.

[@AngleSideSideThm]

LEDs have a minimum turn-on voltage and then after that more voltage = more light, so not entirely analog. There are also alternate dimming methods such as PWM.

[@MrBanzoid]

Slight correction. The cathode in an alkaline battery is manganese dioxide, not magnesium dioxide, which chemically doesn't exist.

[@L1m3r]

Funny and all but soooooo incredibly wrongish / half-truthisch / missing important stuff... I'm waiting for the disclaimer that none of the information presented is researched to any reasonable depth. He's at most skimming the surface of every topic... I'm not a 100% sure but I think one of Konrad Zuse's mechanical computers was/is Turing complete. @38:23 - LEDs aren't digital devices. They aren't just on or off. Limit the power (=current in this case) and you can dim them quite a bit. But that is not how classical dimmers nor how most LED bulbs(!) work. @43:00 - AFAIK you can tell Linux (kernel!) at start time which memory blocks to ignore. I'm sure there are automatism doing that on their own too... @45:57 - Hospitals usually have three different sockets: A) normal. B) UPS. C) UPS+Isolationtranformer+noGFCI Granted, AFAIK the third option is only available in intensive care and operation rooms (the sockets stay powered even if ONE(!) device has a ground fault). @50:10 - The same is true for the opposite: When an power-equivalent number of devices are switched off at the same time it's just as difficult for the power grid. (EDIT: Fixed timestamp by ~-40s) @56:50 - "Bolt of lightning can carry(!) 13*10^6 Volts of electricity(!) across .... miles(!)" Wtf? that makes no physical sense at all. Carrying "Volts of electricity" over a distance is just nonsense.

[@bodgemaster7946]

34:22 - We have retail CPUs that clock to more than 5GHz (and, while my I7 7700k didn’t run at 5GHz ootb, it did run stable at 5GHz for a while).

[@dieSpinnt]

"Digital" is just a word describing that something is taking place in STAGES or can be broken down into individual STEPS. (In our electronic/computer context here. Hint: There are several other meanings of this Latin word). So your statement does not make any sense. Like the "aren't" in your first sentence would make alone ... without context ... make no sense:) A LED, which is a DIODE will of course be a digital circuit element, when it is wired in this way. Man ... vacuum diode tubes, DTL, ECL, TTL, etc. technology ... the basis of our modern world ... where and are all used in digital design of logic gates. [1] bodgemaster: The presented train of thought, the history and the graphics including the science around Moore's "Law" do not speak of short time measurements. We are talking about a steady and stable clock (including for ALL cores in a multi-core design) here. You and others hallucinated that extra condition ("boost") out of nowhere. To be clear, let me do the same kind of cheap trick and I will boast with a processor that runs at 12GHz ... no I won't tell you for how long ... that is my little dirty secret. Good idea and nice try but ... your statement is just a lie. At least you get a *facepalm* for that:) Voltage: The electrical voltage (often simply referred to as voltage) is a fundamental physical variable in electrical engineering and electrodynamics. Its symbol is the U (from the Latin urgere: to press). It is specified in the international system of units in volts (unit symbol: V). To denote a time dependence, the lowercase letter u is used for the instantaneous value of the voltage. In Anglo-Saxon the symbol V is used. The electric voltage U_AB between two points A and B in an electric field with field strength vector E → is defined as a spatial line integral along a fixed path from point A to point B. Yeah ... a little strange how he talked about this. The field with that voltage may exist over the distance of miles ... well, even as this "natural" capacitor. And when the lightning strikes ... the current flows to equalize the electrical voltage ... which practically defines an electrical field and the resulting current. Greetings from Siemens: The load control of three-phase synchronous generators/asynchronous generators can be carried out almost continuously and instantaneously. Reducing three gigawatts across the network for a peak is demanding, but it is possible (depends of course of the overall properties ... but as a house-number we choose that value from the video from 1990). In contrary, the reverse conclusion is a completely different case: Where do you get the power that DOESN'T EXIST from and even the network(capacity) isn't possibly build for? That's your big thinking mistake in this point. BTW are you mad because of something? You don't get yourself everything correct, nitpick like a boss and you demand what? ROTFL Am I in the wrong universe here, have I skipped something or had Dylan claimed that this will be a boring physics lesson? NO! Man, or should I say Karen? Get over it! He is a programmer and a musician. Oh and a lot more funnier than you unavoidable "petty comment columns witch hunters". Sorry for that, but you guys suck!:) [1] By the way, mains-powered LEDs with electric/electronic-ballast devices are something completely different. There are not only a dozen different ways in which dimmer and current control or phase angle control or pulse-pause ratio control (or, or, or ...) can interact, but also countless different designs and implementations. Explaining that, including how LEDs work correctly in detail, would take at least two semesters of electrical engineering. And not an hour-long lecture. What the hell are you expecting here?:)

[@L1m3r]

@@dieSpinnt He IS using DIGITAL in that 1/0 off/on sense in that LED LIGHT-BULB(!!!) sentence thou. And even normal diodes aren't just digital in no shape or form. They are non-linear devices which are often used in digital on/off applications (in the idealized sense). In what way does "LEDs aren't digital devices" not make sense? They are non-linear and their brightness depends on the amperage(?) flowing through them. Btw. as mentioned - the context is LIGHT BULBS & dimmer switches not working together because he claims LEDs are digital devices! Which is just false. What more context do you want me to add? "the current flows to equalize the electrical voltage" ?! It equalizes different electrical potentials. Dunno, maybe my definition of the word(!) Voltage is slightly different but "equalizing voltage"? Maybe when charging large Lithium storage cells in parallel before putting them in series so the active BMS doesn't take weeks to "equalize the voltage" of all cells (as in every cell has the same electrical potential difference = same charge level). You and him both - but him more so - use "Voltage" in a way that makes me cringe. Maybe not technically wrong but still ... As far as I can tell the NDC is a technical conference - his talk is more akin to a comedian but the disclaimer is missing (which is my point). And I saw the clip, noticed a lot of oddities and commented on some of them so other views get a chance to - I don't know - not learn sth. wrongish?

[@L1m3r]

@@dieSpinnt Re: Power Grid - yes, it's different in the sense that a huge sudden demand lowers the frequency of the whole net below 50Hz and the (steam) generators literally turn slower under the load (which is why the frequency drops if I'm not mistaken). A sudden drop in load does the opposite: The load on generators drops, they turn faster (momentum) & the frequency rises above 50Hz (dunno maybe 0,05Hz?). With "the same is true for the opposite" I meant the "one of the biggest problem of managing any kind of grid infrastructure". It may not the same size of problem but the grid will break down (parts switching of / disconnecting / etc) too if the frequency rises too high (solar power stations switching off, etc.).

[@Zlodej5]

BRiliant talker, even if for the first time it was out of his scope and it was noticeable with quite few inacruracies. Nobody is perfect every time.

[@frischifrisch6860]

Thank you for the nice talk. I wish you a sunny spring, best regards and see you soon ( ´◔ ω◔`) ノシ

[@ICountFrom0]

One of my favorite short story fictions involves aliens invading us by breaking electricity. They somehow blanket earth in a field that makes it so that only organic electric actions occure, and they just happen to have organic tech. (insert handwave here). We end up fighting back with mechicanally aimed rockets. Feels inspired by babiage and russia.

[@diablominero]

There's batteries that run on anaerobic bacteria shuttling electrons around to metabolize things. Neuroscientists have developed neural networks that use actual rat neurons to do computation. We might not win against those aliens, but we could do a surprising amount with only organic electronics.

[@diablominero]

Some old pacemakers had plutonium and metal plates to collect the electric charge of the radiation it emitted and ran on that energy for the rest of the person's life without being recharged.

[@davidmiller9485]

Also You are wrong about U.S. voltage. In Residential areas it's 240 to the power box. It's split onto two legs for 120, unless you have a dryer or Ac or other appliance that requires 240. In that case they run both legs to the outlet giving 240. I'm starting to think i should make a list. The majority Like 90% plus of the crude oil in the world, has NOTHING to do with dinosaurs. It does have a lot to do with plants. Ferns mostly. Edit: i have big fingers and didn't look until after i hit send. Fingures.

[@petergerdes1094]

Right, I was here just to make this point. In the us, we use a 240V transformer with a center tap and the customer gets two live connections and a nuetral with the live you nuetral being 120V and live to live being 240V. So we can use lower voltage for most circuits making them safer and more voltage when needed. Though I think some countries actually get 3 phase power to residential customers which has advantages for some applications.

[@traveller23e]

@@petergerdes1094 For the purposes of the consumer though, it's effectively 120. Traditionally it's also called 110, much as 240V in Europe is also called 220 or 230. The margins of error are large enough that it doesn't really matter that much, and in fact you can see in the photo of the voltage set switch in his presentation it's labelled "115".

[@petergerdes1094]

@@traveller23e Many of us have 240V appliances in our homes. I just wired an induction stovetop that way from my bog standard breaker panel.

[@traveller23e]

@@petergerdes1094 That requires a separate 240 v circuit run from the breaker to the kitchen. It's also common for washing machines over there, so I hear. But stick a multimeter in one of your standard electrical outlets and tell me what voltage you get. That's the voltage most appliances (electric toothbrush chargers, computers power supplies, sewing machines, electric mixers, light bulbs...) use and are configured for. Although now it's more common to have devices that can handle either 120 or 240 without a problem, until recently this was fairly rare; about ten years ago I distinctly recall causing all the magic smoke to leak out of an American electric toothbrush charger by plugging it into a european outlet with just an adapter (no transformer).

[@petergerdes1094]

@@traveller23e Yes, I know. I've actually done that wiring. Yes, outlets are 120V but it's not like 240V is irrelevant or invisible to consumer,. Almost every house has at least one 240V circuit in it.

[@GrantDiffey]

ARRRG! that's a TEMS machine not an electroshock therapy machine (Electroshock therapy is much higher voltage applied to the temples as a neurological/psychiatric treatment)

[@HMPerson2]

not TEMS, TENS (Transcutaneous Electrical Nerve Stimulation)

[@GeorgeTsiros]

4:50 "the engineering precision did not exist"... didn't we see the antikythera mechanism in the very previous slide, which was like 2000 years earlier?

[@diablominero]

Tesla coils send arcs up to a few meters, but they send wireless power as radio waves substantially farther than the range of their arcs.

[@maxsnts]

Radio above 2.4Gh? Isn't radio in the MHz range?

[@entcraft44]

The graphic is ordered by wavelength, which is inversely proportional to frequency. So wavelength increases from left to right, and frequency from right to left. You are right about most FM radio being from 30 to 300 MHz.

[@ironnoriboi]

34:25 But the new AMD 7950x has a boost clock of 5.7GHz. That is closer to 6 than to 5. But the base clock is only 4.5GHz, fair.

[@bloody_albatross]

"There's only one type of radiation." Well, particle rays are also counted as radiation (e.g. alpha and beta radiation, cosmic rays, and others) and there are gravitational waves etc. 🤓

[@crides0]

I think the sour taste he mentioned even you mouth strip wires is probably from the metal (ions) itself. Burning is guaranteed. I thought basic stuff tastes like some variation of bitterness (see basic/alkalinic noodles)

[@diablominero]

Lye actually tastes like licking a battery (because taste receptors get overwhelmed by alkalinity and send random noise). If you've ever tried cold-process soap making, you might have done the "zap test" where you lick the soap to see if all the lye has already been consumed or not. Milder bases are often described as "bitter" though that doesn't perfectly match my experience.

[@maxheadrom3088]

Quantum tunneling is when two particles occupy the same place in space at the same time - but since it's quantum and they are waves, what happens is they are two waves in opposite phases. We can buy tunneling diodes on any electronic parts shop.

[@gcinko1525]

~29:40 more voltage doesn't make them faster, just the opposite. The transitions betwéen states can be quicker by lower voltage. Heat is produced during transitions => higher frequency = more heat.

[@jan.tichavsky]

If you can manage the heat then switching with higher voltage is indeed faster and you have more margin to define and correctly recognize the logic levels.

[@gcinko1525]

@@jan.tichavskyFrom Your logic is transition between 0 and 12V a faster than between 0 and 5 or 3V. Any explanation, why?

[@kamilchen4]

6 moths after video published here, we have desktop PCs running 6GHz, amazing world.

[@richardforster1239]

T-shirt at 1:39. A Lazuli fan I see. 👍

[@user-jn4sw3iw4h]

46:00 Then again. If you do something, someone *really* doesn't want you to do. The trip to the hospital afterwards can't possibly be shorter.

[@MrBanzoid]

Curious Marc has uploaded a couple of KZbin videos on the Russian mechanical computer that cosmonauts used to calculate reentry vectors.

[@diablominero]

US houses actually have plenty of access to 240V potential differences. We just create them as pairs of 120V phases 180° apart. So we can still run high-power devices like electric stoves and air conditioners, but there's no wire in the entire house that you can get a 240V shock off of, and that's safer for small children with forks.

[@johanneswerner1140]

Voltage don't kill. It's them Amps. And please install fixtures with some protection. They have been standard here for a couple of decades... Oh and: By the same logic you could say that in Europe we all have 380V everywhere. Technically this is true. I did run those to one basement room, they are not connected to the fuse box yet, but I could run a big ass lathe or mill if I wanted (and get my sparky in to hook it up).

[@diablominero]

@@johanneswerner1140 have you ever heard of an equation called "Ohm's Law"?

[@TobiKellner]

​@@johanneswerner1140 only you don't get amps to flow through a human body without enough volts...

[@entcraft44]

@@johanneswerner1140 Not everyone in Europe gets three phase power, though a lot of us do. Also, at the frequencies and voltages present in your household, voltage is directly proportional to current (Ohm's law). (But beware of dielectric breakdown: The resistance you measure with a cheap multi-meter at e.g. 9V may be much higher than the actual resistance at a hundred volts. Ohm's law is a useful concept, but not a safety rule!)

[@MaverickBlue42]

Technically speaking, the majority of the biomass that died during mass extinctions to later become oil was all plant matter. There's definitely dinosaur bits in there too, but plant biomass outweighed the biomass of all creatures combined well over 100:1, just sayin....

[@AnonymousMaykr]

34:15 yep, he's wrong. The i9-13900K boosts to 5.80 GHz (according to Intel, at least)

[@dieSpinnt]

The presented train of thought, the history and the graphics including the science around Moore's "Law" do not speak of short time measurements. We are talking about a steady and stable clock (including for ALL cores in a multi-core design) here. You and others hallucinated that extra condition ("boost") out of nowhere. To be clear, let me do the same kind of cheap trick and I will boast with a processor that runs at 12GHz ... no I won't tell you for how long ... that is my little dirty secret. Good idea and nice try but ... your statement is just a lie. At least you get a **facepalm** for that:)

[@shnake_case]

@@dieSpinnt dude chill, it's not that serious lol

[@iFireender]

LED's are not digital whatsoever, it's just that they're controlled by current rather than voltage, which turns into somewhat of a problem once you have capacitors and stuff in there ;) They're 'digital' only in a way that they emit light while current is flowing through, and don't when there isn't, compared to incandescant lightbulbs which basically lowpass your signal already because heat.

[@BrendanClements]

Hey, KZbin. All those fancy algorithms and it took you half a decade to show me this guy?

[@diablominero]

LEDs aren't digital. They just have different analog behavior than ohmic devices.

[@dieSpinnt]

"Digital" is just a word describing that something is taking place in STAGES or can be broken down into individual STEPS. (In our electronic/computer context here. Hint: There are several other meanings of this Latin word). So your statement does not make any sense. Like the "aren't" in your first sentence would make alone ... without context ... make no sense:) A LED, which is a DIODE will of course be a digital circuit element, when it is wired in this way. Man ... vacuum diode tubes, DTL, ECL, TTL, etc. technology ... the basis of our modern world ... where and are all used in digital design of logic gates.

[@diablominero]

@@dieSpinnt an incandescent lightbulb can also be driven by a circuit that has a few discrete modes (like a non-dimming lightswitch). But incandescent lightbulbs are are generally considered to be analog rather than digital devices even despite that capability. So it sure seems like you must be misunderstanding what people mean by "digital device"

[@larryfulkerson4505]

Almost everybody knows about Albert Einstein but almost nobody remembers his older brother Frank who had a good run in motion pictures about 1930 or so.

[@matrix01234567899]

LED isn't digital, it san also make various brightness, but it isn't powered directly by ac power but using additional circuts to make ac into dc.

[@dereks6742]

What do the NDC letters stand for?

[@averyalexander2528]

-Double-check me, but I think it's National Developers Conference?- Norwegian Developers Conference

[@CattleRustlerOCN]

I'm from the future, the new speed record for CPUs was set on an Intel 14th gen core i9 at just under 10 GHz with liquid nitrogen cooling. I'm sure AMD will take back the crown at some point soon. And that same Intel processor operates normally at above 5 GHz, at roughly 6 GHz to be fair.

[@matrix01234567899]

34:24 Intel Core i9-13900F was already 5,6GHz when this video was uploaded

[@raydlee.mobile]

Nice cross-mention of @tomscott at 00:22:00 and the parks telescope video - kzbin.info/www/bejne/bKCWaXZjqtGnrNk

[@connorjames1066]

Lmao, watched this after a rant on nuclear, so being from Cali first plant in my head was Diablo

[@mattymerr701]

The latest i9 already boosts to 6GHz

[@blacklion79]

USSR got very high quality photos from Venus. Yes, these apparatus were dissolved at least, but not momentarily.

[@wagyourtai1]

time to ruin things in the youtube comments again. technically america is 240V, we split it in half for most outlets and stuff tho, except for like the laundry and airconditioning and stuff

[@Joshinken]

I love that im watching this presentation claiming we won’t have cpus running at more than 5ghz on a machine that clocks in at over 5ghz and that was already out when this video was uploaded Yeah the newest amd processors clock up to 5.7ghz, and the 7900x im using came out 27/9/2022, BEFORE this talk was held Granted, it don’t run at that speed all the time, but not because it can’t but because it would create too much heat and waste too much power

[@CryptoWulfApp]

Thats a great talk but "Decentralization" is already defined and no buzzword. It means that no entity has the power over a network or can censor it. Every single person defines it own rules in the Bitcoin network and thats not possible for any other. Also: Bitcoin != "crypto"

[@goldnutter412]

Not exactly, core code team plus social consensus mechanism controls the network, miners facilitate it in terms of how much security is created. Not a DAO structure, because it is not autonomous.. such networks inherently have a bootstrapping phase to get anywhere which was Satoshi and Hal Finney basically, and momentum or critical mass is a symptom. An element of groupthink is normal until it is a simple thing everyone understands. We are still far from that point.. Bitcoin is the first cryptocurrency by definition, lets just hope it ends up solidifying where it belongs. It can be an efficiency yardstick for fiat systems to try to keep up with, through DAO structures that take away the complexity and inefficiency of governance by humans. We don't need AI, we need protocols and automation.. complexity management ! half of the things people argue or complain about are non issues. It's the free will thing.. and trust is incompatible with the unknown at a certain point. People feel the need to have control ;) What he's referring to is the nature of open source and free will, and the internet being still the old internet. A level of deception is seemingly easy (or actually easy), in some places people aren't technically breaking laws making coins with 1 trillion, giving out some to certain people etc etc, but it is morally disgusting. It exploits the inability of people to understand what the fully diluted market cap is, regardless of how issuance is done or locks.. DAO's are the only solution to greed and clearly compatible with current governance structures, just a parallel that ends up being far LESS complicated.. once we have some core functional things in place. Human verification fully decentralized with no key storage soon, hardware is hard

[@edgeeffect]

What on earth are you two going on about?

[@goldnutter412]

Nov 2026 it is !

[@Graham_Wideman]

Arrrrg -- the entertainment value of this talk is so badly undermined by all the misinformation and cringey mistakes. From the backwards spectrum (wifi is NOT usefully described as "above" infrared... ) to 2.6 GHz described as 2.6 million, grids described as 110V and 220V when the mains outlet standards are 120 (US) and 230 (EU) (and as many commenters have noted, US uses 240 to the house). And predictably, "put the whole computer in a bath of liquid nitrogen" is complete BS.

[@THB192]

Additionally, cosmic rays are very real and very much do cause memory errors, which are rare but occur more at higher altitudes. In fact, Google observes more crashes on their servers during periods of high sunspot activity because they're operating at such high scale.

[@MartinMaat]

He had to leave his guns outside in the saddle bag, did he?

[@humanperson8363]

Why not just make cpus really big but so you can cool them better, same density just more voltage and more room for cooling

[@pmmeurcatpics]

The very reason why they shrink transistors etc. in the first place is to minimize the distance the current needs to travel. One might think that the speed of light is plenty, but the round trip time between different components of a system does matter! This is why there are different levels of memory: first the registers, then the caches (L1, L2, L3, possibly more?), then SRAM and DRAM, then HDD/SSD - each one bigger but further away from the CPU than the last, with the latter being responsible for the greater latency (which is why low-level programmers generally try to fit as much as possible in L1/L2, for example). Extending the CPU like you suggested would kind of defeat the whole purpose. But this _is_ actually done - that's what cores are! Given, each of those may have their own L1/L2 caches for example, and then have a common L3 all the cores share, and then the RAM, or alternatively only share RAM... - I'm pretty sure this varies across systems

[@ironman5034]

There 5.1ghz

[@BryonLape]

Oil is not a fossil fuel.

[@alexanderkalashnikov2067]

This guy has no fucking idea on what he is talking about.