>DDR5Is it a scam? Why do I hear that it requires 30+ seconds to boot with DDR5? Hard to believe that's true because it's slower than 10 years old PCs.
All technology since 2014 has been a scam
>>100321366thats because windows has to boot on the tpm chip for remote access by feds if you want a real secure machine you should stick with talos II
>>100321366You can disable it in BIOS, but I kept getting random BSODs so I left it on.
>>100321366DDR5 takes longer to train the memory because it's so densly packed on the chipthis is also why it has built in ECC modules, becayse it's much more likely to suffer bitflips so the ECC has to check the data in the DRAM chip to ensure it is still validbut DDR5 is significantly faster than even top end DDR4. DDR5 starts at 5000MTs which is utterly insane, that is faster than most CPUs. also, DDR5 DRAM chips all run in effectively dual-channel mode per chip, meaning each chip can be accessed from two places at once, effectively halving any delay in memory checks.the only downside is the training but that only happens once during boot and in theory shouldn't need to happen again unless you tweak the timings or something. memory training is just the process of organising the chips to run more effectively in your given system.
>>100321557Wait, so it happened once? Everyone says it happens every time you turn on your PC. Or is there a fix for this? I honestly can't comprehend that someone would be ok with such long boot times.
>>100321557Yo momma can also be accessed through two places at once.Other than that, many thanks for the extensive explanation, anon.
>>100321586AMD does some stupid shit on each boot and can take 10 to 20 min to boot.
>>100321586It's not really a problem. I usually just go to the toilet and make some coffee after I boot my PC. It's usually almost done by the time I get back.
>>100321556Disable what? XMP/EXPO?
It’s just some shit motherboards. See the first cheap b650 video from unbox
Yeah, you have to wait an extra minute or two on every boot. But it is blazing fast in use. A fine deal if you ask me.
>>100321908I doubt there's any real life and noticeable difference from DDR4
>>100321858Fast boot
>>100322059It's a regression. The opposite of progress. Except you pay more for that.
>>100322084 You measure techloliygical progress by boot times? χαχχαχαχχαχαχαχχαχχαχα .........
>>100322059It's fucking crazy in 2024 to wait as long on supposedly newer and better tech.
>>100322102You need to have a non-goycattle mindset to understand, don't bother.
>too pussified to reply directlyKek, goyslave.
>>100321586It's supposed to only happen once, but in the world of garbage hardware running ontop of garbage hardware running garbage firmware reality turns out to be different then what was promised and for some people it runs on every boot
>>100321366YesJust use pregabalinOS
>>100321557What do you think the DDR stands for? It is 2.5MT, not 5MT.
>>100321586For me it happens every time I completely unplug my PC (it starts once for about a minute, restarts and then boots normally in about 30s), but after that is done, it only does a normal boot.It still takes about 30s to boot, idk why, but it doesn't bother me much
>>100321366>it requires 30+ seconds to boot with DDR5I have 256GB of registered DDR5 and it takes about 3 seconds to POST, and that's likely only because it's a workstation board (Xeon W)So idk what this nonsense is about
>>100322230Of course it doesn't bother anyone because you can't and don't want to return it and there's no other choice but to use it.
>>100322199no, it's 5000MTs, falsly advertised as 5000MHz. it's more fair to call them 2500MHzif they were actually 5000MHz, it would be 10,000MTs, as there are two transfers per hz not the other way around, which you should know given you expressly called it Double Data Rate.if it was 2500MTs as you say, they would be 1250MHz
>>100321366There's a guy on Level1Tech Forum that claims his 4x32GB DDR5 set boots 20 minutes
>>100322403whats that got to do with DDR5 memory training though
>>100321586It happens to me every reboot, even without poweroff, even if I disable memory training. But maybe that's because I have a MSI Tomahawk X670E. Everyone faints hwne they hear I bought an MSI motherboard
>>100322436So you need to turn on PC, take laxatives, shart and then take a shower. By that time everything will boot and there wouldn't be a problem.
>>100322449MSI is bottom of the chinesium barrel chinkshit, costs far too much for what's basically a no-name brand synonymous with DONHUA JINGLONG FACTORY GOOD BEST PC GAME EPIC MOTHERBOARD FOR HIGH SPEED AND BEST PERFORMANCE listings on aliexpress
>>100322470At least it has plenty of PCIE slots. Yes, I need all 4 of them
>>100321366ddr4 chads win again
>>100322479>4 is a lor for an x-series chipsetbro...
>>100322501Appreciate, but not buying Intel
>>100322470Is Asrock better? It funny, because I've heard so much shit about Gigabyte and Asus.
>>100322479>>100322501sorry thay's not fair, x299 is a workstation chipaet, gigabyte doesn't have any x670- OH WAIT THEY DOMSICHINKS ON SUICIDE WATCH
>>100322522I'm upgrading my wife's rig from Asrock B550 Phantom Gaming 4. Compared to the MSI X670E, it's rock solid and blazing fast
>>100322530What board is this?
>>100322535Thanks. I'll get Asrock for my wife's son.
>>100322522asrock are ok but don't fall for their lowend shit. they do good WS stuff and asrockrack make great server boardsasus is a gamer meme, overpriced but it's at least passable. they used to be a lot betterz these days they just make flashy shit and charge a lot more for the same tier of product because "it has gamer lights" or "it has matte black everywhere"gigabyte is where it's at
>>100322553May I ask what exactly is a base of your judgement? Brand loyalty or personal experience?
>>100322572i build PCs for a living and provide warranty support, so I see all the faults. MSI is by far the worst with the most RMAs and misc faultsgigabyte has given me the least trouble and are usually some of the cheapest to order, and since i don't price differently for vranding i make the most margin lol but seriously it's the least amount of issues. Asus is also fine mostly but just overpriced. asrock i onky really use their high end stuff eg asrockrack, their consumer stuff never has the right features I need.i also use gigabyte personally in my desktops and servers and haven't had any major issues except one DoA board. Asus is also fine mostly but just overpriced.
>>100321908>Yeah, you have to wait an extra minute or two on every boot. But it is blazing fast in use. A fine deal if you ask me.sounds like Windows7
>>100322613I'll keep your info in mind for my next build (it will be soon). Thanks.
>>100321366Works on my machine. 5 second boot time with a 7800x3d and PCIe5 SSD
>>100322754Which mobo?
>>100322758ROG Strix B650E-E
>>100321366>Why do I hear that it requires 30+ seconds to boot with DDR5?Over a minute here.Boot time absolutely sucks, yes.It's just the POST though, booting the OS is super fast.
>>100322436>4x32GB DDR5 set boots 20 minutesdamn this is so funnydropping $2k on a fresh new pc and takes 20 minutes to boot
>>100322436>Falling for le bigger memory=gooder memeYou don't need more memory, you need faster memory. Casuals don't seem to understand this.
>>100322059>why would anyone sane give a fuck about 30+ second boot time?that's the spirit, bloated apps causing wait times are not a problem either. It gives time for reflection.Seems everyone is a zoomer these days, no matter their age.
>>100321366Wrong, kid. Memory training has a been thing since Sandy Bridge/Zen. It always take time to do. It just impatient spastics that are making a fuss over it. >>100321557All factory overclock DIMMs take longer to train. The gap has grown with DDR5 because they are pushing motherboards and controllers to their limit.
i fucking hate this shit so much>>100321586amd has something called memory context restorethis should, in theory work automatically i.e. the first couple of boots are slow to train, then it remember the profile and boots are fasterin practice, with my mobo, this setting is worthless when set to autoi have to manually force it onbut if you change your hardware and forget to then turn it off, you'll keep bsod'ing (i hate to reset cmos once while learning this the hard way) because the old training profile is no longer valid and needs retraining
>>100322059Exactly, it is a nothingburger for 99% of the userbase. They rarely do a hard reboot and just leave their systems in sleep/hibernate.
>>100322199The number stands for generation of the standard.
>>100322944>Wrong, kidYes, there's no difference between my old ass PC booting into an OS in less than 10 seconds and 30+ seconds on a new PC that is three-four times more expensive.
>>100321366>>100321557>>100322944>>100322981If JEDEC isn't good enough for you then you have zero reason to fuss about training timesThe world doesn't revolve around a few autists with their toy PCs
>>100322449I have an ASUS ProArt X670E with 64GB of RAM.It's also slow: about a minute to POST, every time.Its not a huge issue but it's annoying enough I leave my PC running 24/7
>>100322997JEDEC is retardedly slow for DDR5.We're not talking a few % here.And all the manufacturer benchmarks show "overclocked" RAM, you're paying for the higher speeds.
>30sec boot? that doesn't even happen>ok but it only happens once>fine but it only happens if you unplug it>alright it always happens but it's only 30 secs>it adds 1-2 mins to boot but i can just go grab drink meanwhile>ffs who even cares about boot time?1?!?!Every single time
>>100322981That decade old system is a lot slower at a number of tasks especially stuff that is demanding. The whole memory training is a nothing burger. It only does a full season once after memory configuration. Seqenstial trainings are much quicker. The said system still boots lot faster then system that are equipped HDDs and HBAs were commonplace.
>>100323048Then don't bitch then if you feel it so importantThe ironic thing is I bet most of the tards complaining don't even have a 4090 or even play their gaymes at 1080p where it matters, but bigger number better wins out against everything
>>100322436Because the retard is absolutely taxing the shit of their memory controller/motherboard. I suspect his setup is barely stable and most of that boot time is just system trying to training again and again to find moment where it is briefly "stable". In the past, his BIOS/UEFI would have just given him the "memory error" on boot-up.
>>100322530>we will never know what board is thisok keep your secrets fag
>>100322873>just the POST though, booting the OS is super fasHe claims it's a 6000MT CL30 memory
>>100322059my computer with 2GB of DDR2 ram in 2005 booted in 8s
>>100323202This is only made possible with modern SSD media which at the time that platform was relatively new was only exclusive to enterprise world. It operated with HDDs that took a while to load-up the OS environment.Modern systems are just as fast after a full memory training is done and thjey aren't running factory overclocked DIMMs that push their motherboard/controller to their limits. Silicon lottery for memory controller is becoming more of a thing now.
My 300 € laptop from 2014 takes 4 seconds from fully shut down to the login screen, how are you niggers taking 7-30 times that amount and being fine with it?
Are DDR6 leaks about two hour boot times true? I guess we'll have to convince ourselves that's normal.
>>100321366DDR5 on AYYMD is like that. On Intel it is also bad but we’re more focused with processor issues over here
>>100323008Intel or AMD?
>>100323348that's nice dear, but we're talking about DDR5, which due to its' more complex nature, requires a training cycle on first boot, or at least it should only affect the first boot
>>100323348>are you niggers taking 7-30 times that amount and being fine with it?It's retarded overclockers not understanding that DDR5 is nearing the limit of what's possible on DIMMs and being on the edge of stability means shit has to retrain
>>100323400AMD
>>100323130Yeah I don't know why that would be then>>100323398>DDR5 on AYYMD is like thatWorks on my machine. Boot time is pretty much instant.
>>100323425i personally have bothered to look into avoiding training on boot, since i only reboot my machine about once a month for kernel updates
>>100323448Same but with a ProArt 570X
>>100323413 & >>100323425See:>>100323070
>>100321557>5000MTs>that's faster than most CPUsUhhhh, no, it's not...
>>100322873I do rendering, therefore I do need more memory you fillthy casual.
Brainleto here. This seems like the place to ask:Will a user ever feel the difference between 8000 MHz DDR5 and something less like 4800 MHz? Assuming everything is installed correctly and in a correct configuartion.
Stop buying shit mobos and get an Asrock.
>>100322354Who is falsely advertising it? Its only a colloquial issue that people confuse MT/s for true clock rate.
>>100323452That’s good. I think mine hovers around 13.6 seconds
>>100323955 I can tell a difference between 4800 and 6600mhz but there’s no xmp profile for 6600 which makes using it mildly annoying
>>100321462fpbp. underrated. based.
HAHAHAHAHAHAHAHAHAH I TOLD YOU GUYS THERE WAS NO HOPE! I told you guys that technology peaked 5-10 years ago and everything we're seeing now is just desperate copium from hardware corporations. HAHAHAHAHAHA, I'll enjoy my 5 year old PC that A) runs every softwareB) is stableC) hardware support up the wazooIt feels good to be proven right.
>>100321366works fine on my machine. 6.27 seconds
this is a king of manlets kind of moment for meI have an SSD raid 0 for wangblows, no lessand that fucker used to startup real quick, I'd say like 10 seconds or so, now it takes like 1.5 minutes if not morewangblows, not even once
20 seconds at 32GB dual channelAMD ASRock livemixer
>>100321908what? no. i just put together a ddr5/am5 build together 2 months ago and i've only had that 'long boot' occur twice
>>1003236642.5GHz is faster than most CPUs you can currently buy
DDR5 has been super unstable for me so far, although BIOS updates within the last year have seemed to have helped.>buy 6000MHz kit>turn on XMP>post times 30-40 seconds EVERY boot>turn on memory context restore>no difference>reset to stock 4800MHz >5 second post timesNever had any of these problems on DDR3. It just werked. Both DDR4 and DDR5 have been a nightmare for stability in my experience
1) flash newest uefi bios2) enable memory power down3) enable memory ontext restore4) enjoy 12sec boot from biosstupid goys
>>100329404Have you noticed a difference outside of synthetic benchmarks? Just run it at JEDEC speeds.
>>100321462Do you recall where you learned that?
Fast boot a shit
>>100331078lmao
>spend thousands on a DDR5 system with 8 channel memory >still dont get better ai performance than adding another GPU>ddr4 is fucking fine for 99% of tasks>just double it if you need more.who is this shit for?
>>100331425GPUs cost thousands...
I've been wanting to experience the latest nvme drives in RAID0 with coreboot enabled and 2 sticks of 48GB DDR5 to see how fast it can boot but this thread makes it sound like it's awful
>>100331699so does a CPU that can do 8 channel on DDR5 platforms.
>>100331787Yeah but then you have a fast CPU instead of a GPU that needs all this extra weird crap.
>>100331865>but then you have a fast CPU that still cant outperform a gpu at 1/2 the price or worse.
>>100323070Once its trained it's fast. Besides how many times are you starting your fucking computer? Use sleep mode you dumb cunt.My computer gets a reboot like once a month.
>>100321557>faster than CPUcache-let retard alert
>>100323070Why are you rebooting a machine like this often? Did you put fucking Windows on it?
>>100332047>once its trainedthis faggot thinks his ram is running shitty python machine learning algorithmgood morning saar
>>100321366>Why do I hear that it requires 30+ seconds to boot with DDR5?Memory training, though most boards should have a setting to remember the last training results so it doesn't need to run every time.
>>100331078Framerates in some games are a bit higher, but generally no, there's not a huge difference in everyday use. If you did a blind test without framerate numbers, I doubt you'd be able to tell the difference.
>he didn't get holy grail 10th gen intel pcie 4 ddr 4 monolithic fine winesaw this coming a mile away
>>100331277lol its insane that people still act like this outside of reddit
>>100331960cpus aren't competing with gpus
>>100332162https://vtechinsider.com/what-is-ddr5-memory-training/
>>100322945>amd has something called memory context restorethanks, firmware time went from about a minute to 12 secondsthis is good enough for me
>>100332692theres nothing cpus do that require ddr5 in my systems nor the systems of most home users
>>100332619>He fell into 14nm+++++++ meme
>>100332689>asking for evidence: LE BAD!
>>100332847are you braindead?that's the point
>>100332721the retarded chatbot that generated this text can't even keep DDR4 and DDR5 straight
>>100321366What the hell is all this about?Does it affect laptops too?
Considering that we're about to get new 5000 series cpus, there's little reason to move off of DDR4 for now
>>100322551Thats what I get the bulls when they come over
>>100332908the problem comes from asking a source on something that's intuitively obvious.
>>100321366Memory training has always been a thing. Even on my core 2 duo rig with DDR2 ram.Admittedly it takes way longer on DDR5 ram.
>>100328659yeah but they all boost up to like 6 million ghz anyway
>>100321366boots after hardware change do take long, i do not notice anything unusual in ddr5
>>100321586>Everyone saysYou should check what motherboards they have, for years now MSI have taken like 40 seconds to boot on amd it's a problem that only happens with MSI boards. I have stopped buying them
>>100321366Dont careStill sticking with 32gb ddr3
>>100321586happens everytime the cmos is cleared or it when detects an error during initialisation
Holy shit, my ancient 4690k with DDR3 and sata ssd loads win10 faster than my another pc based on ryzen 5800x with DDR4 and nvme.Why do they do it ti us?
>>100336720no, it loads the OS much fasterit takes longer to POST because you have retarded UEFI settings and your memory has to retrain every POSTafter the UEFI logo flashes om the screen, your OS is booting - i gauruntee that time takes almost nothingbefore the UEFI logo flash, it is doing POST, if that takes a long time it is 99% of the time a memory issue with shitty auto OC settings changing every timelearn how your computer works and it'll work much better for you, or just buy a prebuilt and blame them when it fucks up
>>100336756I am fine with my setup. It takes 11.4 seconds to load into OS and start to shit-post on this pathetic image-board infested with arrogant trannies like you and stinky poojeets like OP.
>>100321366>he didn't buy gigabyte
>>100336989>23 secondsMy laptop from 20 years ago used to boot faster than this. What's the point of new technology if the PC doesn't boot in 1 second?
>>100321366I run DDR5 with EXPO/DOCP, and it feels like I've gone back to the 90's. I'll turn on my computer then walk away to do something else while I was for it to boot. I'm not entirely sure if it's the EXPO profile, MOBO, or the RAM itself.
>>100337308There being such a difference between motherboard manufacturers means it is their fault. I doubt gigabyte which seems to be the fastest has found the perfect solution either.
>>100324306> I can tell a differencehow? you stick your little dick in the DIMM slot? You're a liar
>>100329404>both DDR4 and DDR4don't lump DDR4 in there with your bullshit, faggot. It runs fine with none of these issues.
>>100321366Imagine powering your machines down
>OMG 30 second boot time that you do once in a blue fucking moonGrow the fuck up you retards. Is your life really this boring that you have to find stupid shit like this to cry over on a chinese basket weaving website?
>>100338155Yes, we are not poor
>>100333225What? Sauce nao
>>100338155I expected computers to become faster over time.
>>100338155>once in a blue fucking moonIt isn't though. It's every boot
>>100336373second post best post
>>100338190not happening with TPM and other new shit coming.You can thank Microsoft, AMD, IBM and Intel for this
>>100321557>but DDR5 is significantly faster than even top end DDR4>DDR4 3600 CL14 - 7.7 ns latency>DDR5 5600 CL28 - 10 ns latency>DDR5 8400 CL40 - 9.5 ns latency>significantly faster than even top end DDR4lol, lmao even
>>100338224latency and speed are different conceptsif DDR5 can do a task in less than half the time (higher clock and all chips are dual-access) but takes 20% longer to start, it still takes less time overall and is therefore fastera few MS more latency is irrelevant when it's more than twice as fast at transferring data
>>100338224Yes, DDR5 is a lot faster than DDR4. Learn how memory actually works. Whatever your use case benefits from the extra bandwidth is another matter. Besides, the fat caches on modern CPUs masks almost all of the tangible benefits from tighter timings.
>>100338190They are faster, however the problem is that for most causal use cases we have been scaling hard against the asymptotic wall of diminishing returns.
>>100338336depends on your gayman.If you buy DDR5 because you want more fps and you don't play in 720p you will see no net gain anyway
>>100337308It is the motherboard and memory controller on the CPU. That portion is starting become more like rolling the silicon lottery for factory overclocked SKUs.
>>100322238It is also that you are running JEDEC-spec DIMMs and motherboard that is overengineered to handle JEDEC-spec grades without no hassle.All of the long memory training are from operating factory overclocked memory on boards that can barely handle it.
>>100322449really? I got an msi board. No problems so far.
>>100323413Sounds like the hardware wasn't developed well and rushed out the door. The training sounds more like a bandaid for it's deficiencies. At least we got dumb gamer consumer types who are willing to beta test this stuff at their own expense.
>>100338511eh, i didn't know about it when i upgraded to ddr5. after the training it posts in about the same amount of time as my previous board
>>100338432>All of the long memory training are from operating factory overclocked memory on boards that can barely handle it.It's more the CPUs can barely handle it>>100338511The tech is fineThe entire problem is that you have CPUs designed to handle a certain memory frequency and gaymers and PC tweakers are demanding more than what they are designed to handle.
>>100338540It's more the CPUs can barely handle it. Nah, it is the memory controller then motherboard itself. The DIMMs are rarely the issue.
>>100338386>implying games are the only resource heavy thing people do with their computers
>>100338540Then perhaps they should stop pandering to the gamer types
>Ryzen 7 5800X3D>64GB DDR4 32000 lol the only thing upgrading is my mountain gear :)
>>100338581Making the gamers pc take long time to boot is opposite of pandering them. The guy with xeon system that boots fast is the one being being pandered.
>>100338562no one else buys Dominator 7200Cl24 RGB gayman dragon GOLD edition DDR5 Ram.Maybe some freaks on /g/ so they can create and decode AI child porn
>>100322436does he run like 64 cores too?
>>100338562Yes, let's see this "resource heavy" thing that is made so much better with faster than JEDEC RAM
>>100321366i did a test, got all system go buzzer at 7s, 40s to login screen, i like ddr5
>>100338841how slow is your storage/os that it takes 33s to load?
>>100338399>It is the motherboard and memory controller on the CPU.Wow im so glad amd brought us modular chipsets so they could keep shipping 14nm garbage and pretending it's a good thing.
>>100338193He's saying for a process you do every once in a while. How often do you honestly restart/turn off your pc?
>>100339200hdparm -t says 533-536MB/s range
>>100339480>LinuxThat explains alot
>>100339480sata?>>100339489not really, i'm >>100339200my os takes just under 9 seconds to load up, and it's far from a minimal setup
so this is an amd only problem isint it?
From what I understand the long boot is only on the first install, right?
>>100339506
>>100339773happened again when switched display from integrated to pcie at bios
>>100339773>From what I understand the long boot is only on the first install, right?It's supposed to be, but for some lucky people it's every single boot
>>100321366because DDR5 is unreliable garbage to make line go up. DDR4 was already at the edge.Your CPU's memory controller, as a precaution is basically doing link training instead of relying on cached results, probably because of a firmware configuration that's defaulted to off because of what I said.anyhow, I can't wait til the "rowhammer" resistant shit comes out and itroduces data dependent pauses in memory access. can't wait for the next level fucking spectre and watch as half the vendors fuck it up anyhow.FUCK DRAM.GIVE ME FUCKING MASSIVE SRAM ALREADY
>>100321858memory training
>>100321858
new zen will fix everything.
DDR5 will continue being a meme(for nearly everything) for the next two years at least. Probably more.
>>100321796Imagine living like this
>>100333314Someone here told me CPU and GPU companies started raising TDP to milk extra performance outta their hardware (vs real innovation like in the past). If I recall correctly, he said anything after 12th gen Intel or Fermi (or Pascal) is a scam. Might that happen to line up with what you know?
>>100339624Im going to take the lack of responses to this post as a yes
>>100321366Because it takes time to 'load' the RAM or some shit.DDR5 is the Snap of RAM.
16GB of DDR3REE reporting in
>is [current gen ram] a scam?>yes, because it's new and NEW BAD OLD GOOD NEW BAD I can't afford new so new BADEveyr time.
>>100338562Games are by FAR the most memory-sensitive applications that the average desktop user will ever run. And the reality is that DDR5 makes close to fuck all difference, despite all the hype. Not on Intel platforms where there's a direct 1:1 comparison, and not on AMD where the 5800X3D is barely slower than the 7800X3D (for gaming).
>>100345115>the 5800X3D is barely slower than the 7800X3D (for gaming)This is completely wrong, if you look at games that actually care about the CPU then the 5800X3D gets matched by the 7600 while the 7800X3D is far ahead. Try looking at UE5 games or at least comparisons using an overpowered graphics card at 1080p to remove a GPU bottleneck.
>>100322981>10 secondspoorfag
>>100345157If you have to contrive a scenario with a 4090 at 1080p, than RAM ostensibly doesn't mean dick.This is the most craziest fucking thing about gaymer obsession with RAM and CPU. If you have south of a 4090, the actual consequences of RAM or CPU just got that much tinier.
>>100331425>DDR4 caps out at a pitiful 3200Mhz on a high end 13th gen cpu>despite DDR4 supporting upwards of 5600Mhz>DDR5 support for the same cpu is over 6000Mhz>all of a sudden every soituber calls it "megatransfers" instead of megahertz for some reason
>>100338258That's not how memory is used most of the time. Latency matters a lot.
>>100333314>muh opinion is inherently obvious and true because I believe it>evidence and supporting data is le reddit
>>100333314>intuitively obviousThere is no such thing.
can a n100 or n305 that uses a single stick of ddr5 use it as "dual channel"??
>>100346591DDR5 is only "dual channel" in that it splits the 64bit wide memory bus of a single DIMM into 2 32bit busses.This does give the advantage that the two halves of a DIMM can be accessed in a asynchronous fashion, decreasing latency but it's not real dual channel but a single 4800mhz DDR5 DIMM is still similar in performance to a dual channel 2400mhz DDR4 setup
>>100343171While Intel has indeed massively ramped up its power usage in an effort to remain competitive with AMD, the same is not really true for Nvidia or AMD. Both companies most recent chips will easily outperform their predecessor on the same power budget.
>>100338125Obviously you weren't building computers around the start of the DDR4 era, because your post makes zero sense. DDR4 was plagued with ton of the same stability issues that DDR5 had in its first year or so.
It's over
>>100321366>Why do I hear that it requires 30+ seconds to boot with DDR5?Because you're schizophrenic and haven't been taking your medicine.
