The Critique of Overkill: A Look at Modern Motherboards

In today's world of high-performance computing, motherboards have become a benchmark of excellence for enthusiasts and professionals alike. However, not all motherboards are created equal. In this article, we'll take a closer look at the design decisions behind modern motherboards and examine whether they're truly necessary.

A Look at a Typical Motherboard

Let's take a closer look at a typical motherboard that's currently on the market. As our host points out, it has four RAM slots, but only two are actually needed. This is because modern CPUs have a much higher memory bandwidth than they used to, making two slots sufficient for most users. Additionally, there's a PCI Express slot with SLI enabled, which is rarely used in everyday applications. It's clear that the design team behind this motherboard has made some questionable decisions.

The Reality Behind Motherboard Design

So why do modern motherboards come equipped with so many features? Our host suggests that it's because nobody actually knows what they're doing. They may think they're providing a better user experience, but in reality, they're just creating a complex and bloated system that's difficult to manage. This is where our host comes in – an expert in computer hardware who's here to set the record straight.

Understanding Voltage Regulation Modules

One of the most critical components on any motherboard is the voltage regulation module (VRM). The VRM is responsible for stepping down the 12V power supply from the power supply unit (PSU) and providing a stable 1.2V output to the CPU. Our host explains that the basic circuit consists of two MOSFETs, an inductor, and a diode. When the MOSFETs switch on, they charge the inductor, converting electricity into a magnetic field. The voltage of the CPU is dependent on how long the switch is closed, but when it's opened, the voltage begins to drop.

The Importance of Diodes

The diode plays a critical role in the VRM circuit. When both MOSFETs are open and the inductor is charging, the magnetic field collapses, causing the CPU to overheat. This is where the diode comes in – it's highly inefficient, but essential for preventing damage to the CPU. Our host explains that if the second switch closes too slowly, the voltage can bounce around, causing instability and spikes.

Removing Instability with MOSFET Switching Speed

So how do we remove the instability caused by the bouncing voltage? Our host suggests increasing the speed at which the MOSFETs switch on and off. This is normally done on motherboards, but our host notes that it's not practical to try to switch them on and off at an extremely high rate – every time they switch, they generate heat, which can cause them to fail. Instead, we need to increase the number of MOSFETs to provide a more stable voltage.

The Benefits of Multiple Phases

Another feature on modern motherboards is the use of multiple phases in the VRM circuit. This allows for a more efficient and stable power supply, but it also introduces new challenges. Our host explains that by adding another phase, we can roughly double the cleanliness of our power supply. The number of phases in your motherboard is normally shown as a number like "eight plus two," which means eight phases for the CPU and two dedicated for the RAM.

The Trade-Offs of Motherboard Design

As we've seen, modern motherboards come with many features that seem beneficial at first, but ultimately lead to complexity and cost. Our host notes that there's a trade-off between having more phases with higher-quality components and having fewer phases with lower-quality components. High-end motherboards typically have the most phases and highest efficiency components, which means they're capable of delivering very low temperatures. However, these boards also tend to be more expensive and have better VRM heat sinks – purely for aesthetic reasons.

How Does the CPU Talk to the Rest of the System?

Once the CPU has power, how does it communicate with the rest of the system? Our host explains that electricity signals can travel a long distance in a very short amount of time. In fact, our host notes that at 5,000 megahertz, an electric signal can travel almost instantaneously – but our RAM makers have lied to us about its actual speed.

The Truth About RAM Speed

So why do our RAM sticks claim to be running at 5,000 megahertz when in reality they're not? Our host explains that the data is actually being sent every 0.1 microseconds, which is much faster than the advertised speed. This means that the actual bandwidth of modern CPUs is much higher than we think – and it's all thanks to advancements in memory technology.

In conclusion, modern motherboards are a complex and often overkill solution for everyday users. While they may seem impressive at first glance, they can ultimately lead to cost and complexity. By understanding the inner workings of these boards and making informed decisions about design choices, we can create more efficient and affordable systems that meet our needs without breaking the bank.

"WEBVTTKind: captionsLanguage: en- This heatsink right here, pointless.Four Ram slots, you only need two of them.And why is there a PCIexpress slot that has SLI?No one uses it.Why the heck doeseverybody want all of this?It's because nobody actuallyknows what motherboards areand how they work.So here it is,to the best of our ability,how motherboards work.So buckle up nerds, Linus is on vacationand we're not holding back.Just like I can't hold it backthis segue to our sponsor,I fix it.You like to repair yourown electronics insteadof having to spend hundredson expensive replacementor a pair of services, learn more aboutI fix it's essential electronics toolkitat the end of this video.(upbeat music)- Computer power supplies output 12 volts.If you apply 12 volts to a CPUit looks something like this.Suboptimal, for sure.(chuckles) Suboptimal for sure.To avoid the magic smoke,we need to take that 12 voltsand step it down to somethingmore like 1.2 volts.And this is exactly the jobof voltage regulation modulesthe VRMs that overClockerstalk about so much.The basic circuit consists of two MOSFETswhich in this situation arebasically just fancy switches,an inductor and a diode.This first switch closes whichthen charges this inductorconverting the electricityinto a magnetic field.The voltage of the CPU getsdepends on how long the switch is closedbut the instant it is opened,the voltage to the CPUbegins to drop kinda like a drain.So this is where that diodehere is very important.If both switches are openwhile the inductor is chargedit's magnetic field willcollapse and pop goes your CPUbut the diode is highly inefficient.So it's important to get thesecond switch closed as soonas possible to avoid creatingany unnecessary heat.If you're paying attentionyou might have noticeda problem here though.Although the voltage will be bouncingaround what the CPU is asking for.There are a lot of spikes,not great for stabilityand there are two ways to remove this.The first is to just increasehow fast the MOSFET switcheson motherboards.This is normally done atabout 300,000 times per secondbut even then the voltagecan fluctuate morethan the CPU wants.And it isn't practical to tryswitching a whole lot faster than that.Every single time theMOSFET switches on or offit generates a bit of heatand above about 150 degrees Celsius.The MOSFET's gonna die.So if we want cleaner power,we don't need faster MOSFETs,we need more of them.The circuit we have hererepresents a single phase sun.Another term that overclockerslove to use so much.So by adding another phase, we can roughlydouble how clean ourpower is with the benefitsof additional phases,scaling, roughly linearly.From there, the number of phasesin your motherboard is normally shownas a number like eight plustwo, which means eight phasesfor the CPU and two dedicated for the RAM.Let me put this away.Multiple phases also have another benefit.Say you're a CPU requiresa hundred amps to runwith a single phase.All 100 amps would have to go directlythrough those components.But with two phases, only 50amps to go through each phasemeaning lower rated and thuscheaper components can be used.You might be tempted to think then,well more phases equals more better,which is true to a point.However, as you add more phasescontrolling them can get moredifficult, which translatesinto the VRM as being less able to respondto changes in voltage.Like everything then there are trade-offsbetween having less phaseswith higher quality componentsor more phases withlower quality components.This does create a ratherodd situation though.High-end motherboardsnormally have the most phasesand the highest efficiency componentswhich means the VRM temperaturesshould be really lowbut these motherboards alsohave the best VRM heat sinks.Why purely to look cool.Mid range boards mightactually need good VRM coolingbut for anything otherthan extreme overclockingthe high-end board would probably be finewith no VRM heat sinks at all.It's similar to how you'dbe fine without a shirtfrom LTD store.com, butdefinitely look cooler with one.- Now that the CPU has powerhow does it talk tothe rest of the system?This copper wire representshow far an electricsignal can travel in a nanosecond falls.At 5,000 megahertz your Ram can send dataevery 0.2 nanoseconds in that amountof time and electric signal can travel.This far.So then thisis physically impossible.Ram makers have lied to you,although on your Ram stickin the bios and in windows,your Ram will say it's runningat 5,000 megahertz or whatever.In reality, it's completing5,000 mega \"transfers\" per second.What's the difference?Well, since it's DDR ora double data rate Ramtransfers are done on both the beginningand the end of each individual Hertz.This means the sticks are actually runningat 2,500 megahertz, which is whyif you've ever lookedin CPU Z, the Ram showsup as half of whateverything else claims it is.Depending on what you wantto do with your motherboardknowing how the CPU and theRam connect can be important.Like have you ever wonderedwhy the manual alwayssays to put the Ramin the second and the fourthslots on your motherboard?For this example, we'll be lookingat a CPU with dual Ram channels.This is the most commonconfiguration for consumer CPU'sbut the same basic principles applyto workstations or serverprocessors with more channels.The absolute best case scenariofor a dual channel CPU is tosimply have two ram slots.PCB traces go directly from the controllerto the Ram stick withnothing else in the way.This is why Asus top tier gaming boardsonly have dual slots.And also why some ITX motherboardslike this one can punch way abovetheir weight class when overclocking Ram.Moving onto other boardswith four Ram slots,here you have two main optionsT topology and Daisy chaining.T typology is when thetraces to each Ram stickon a channel are the same length.Which can be great forrunning all four slotsat relatively high speeds.But if pure speed is what you'reafter then Daisy chaining is the way to gowith a Daisy chain motherboard,the traces basically just goto one slot and thencontinue onto the next one.This can cause the speeds to be lower.When all four slots areused, the timing differencesbetween the two stickshas to be figured outby the controller, but it alsomeans with only two sticksyou can run them at nearly the same speedas if there are only two slots.Now it is important to usethe correct slots though.If an electric signal goes through a wireand the wire suddenly endswith nothing connectedto it like this, the signalcan reflect back creating noisein the circuit.By putting the Ram sticksand slots two and fourthe traces and at theRam stick, reducing noisein the circuit andallowing for higher speeds.So basically before you buy Ram ,check if your motherboard has Daisy chainor T topology since it determinesif you should get twohigher capacity sticksor four lower capacityones with higher capacitydims getting cheaper andcheaper, it now would make sensefor most gaming motherboardsto only have twoRam slots optimized togo as fast as possible.Our contact at ASUS agreedbut said way too many people will complainthat they can't upgradetheir memory in the future.Even if statistically very fewpeople actually will do that.But even withouttypologies and slots, likeif you had the CPU right next to the Ramand used gold traces orwhatever for the perfect signalthere would still be somethingelse holding you back,the memory controller on the CPU.Everyone has accepted at this point,there's a Silicon lotteryand some CPU's will naturallybe able to clock higherat lower voltages, but the same appliesto the memory controller on the CPU.Your luck in the Siliconlottery can sometimeshave the biggest effecton how fast your Ram runs.- The CPU and Ram are now workingbut we still need to beable to connect thingslike the GPU, storage, and peripherals.In the past this will bedone through the chip setwith the North bridge takingcare of PCIE and memorywill a Southbridge handleIO storage, audio USByou know, on modernmotherboards though, the memoryand some PCIE lanes areconnected directly to the CPU.Well, chip set handles everything elsewhich simplifies the layoutand allows for lower latencies.Losing a whole chip isbig, but what's even moreshocking is just how fast PCIE has becomewith PCIE 4.0 a one-lane slot can doup to two gigabytes per second.And a 16 lane slot is ableto transfer a staggering 32gigabytes. Second, that's closeto the speeds of DDR four Ram.How, how the heck can they pull this off?Simple, very expensive PCBspreviously was possiblefor lower end motherboardsheavily like four or so layers.But now in order to get PCI4.0 levels of signal qualityeight to 12 layer PCBs arebasically a requirement.Given every two layerscan come with a 20 to30% cost increase for thePCB, the insane pricesof modern motherboards startto make a bit more sense.Even if the majority ofusers will struggle tosaturate a PCI 3.0connection, let alone 4.0so you might have noticeda bit of a theme here.One of the most difficult partsof designing a motherboard isn'tcreating the best possible productbut carefully balancing the actual valueand perceived value likethis motherboard righthere supports SLI hasmassive VRM heat sinks.It has four Demsbecause those are thingsthat people expect.It probably would be betterfor the majority of users.If there were just two ram slotscloser to the CPU, one PCIEconnector, and you knowthe price of this one down here was turnedinto better traces and okay.Okay. The VRM heat sink,it looks pretty cool.And I wouldn't want to give it upbut at least I know that that'sa trade off that I'm making.And hopefully in the futurewe can be more open tocrazy motherboard designsthat either deliver higherperformance or lower cost.Huge, thanks to build soilfor creating the videos usedas a reference for theVRM portion of this video.I've got those link belowif you want to get even morein-depth than we did here.And also thanks to JJ fromASUS for offering some insightthat only a motherboard expert can.Also let us know down in the comments,if you guys want to see more\"turbo nerd\" edition videosand make sure to hit likethis video is only able to godown a couple of rabbit holesand there are so many more here.There are also so manymore high quality seguesto our sponsor.I fix it.Thank you to for sponsoringtoday's video there, I fix it.Essential electronicstoolkit is a great basic kitfor new users.It gives you what you needfor the most essential electronic repairs.It has a compact size includes the mostpopular precision bits,and it's held togetherwith a high densityfoam that you can throwaround without any ofthe bits falling out.It also comes with a lifetime warrantyand I fix it has a bunchof awesome videos toshow you how to, you know, takeapart your device and stuff.Get it today.I fixed it.com/ltt.If you like this video, maybecheck out our recent videoon the semiconductor shortagethat is making GPUS nearlyimpossible to buy for a reasonable price.At least it won't help you buy a GPUbut at least it lets yougo down that rabbit hole.\n"