https://www.tweaktown.com/news/101262/intels-new-core-ultra-9-285k-loses-to-14900k-13900k-amd-ryzen-9950x-7800x3d-in-gaming/index.html
>>102954546Is this a cuck core problem?
>>102954546It's okay, we will just ban TSMC so everyone has to use Intel fabs.
>>102954546>cpu underperforms>cpus and boards get price drops>this is somehow a bad thingLet me guess it's only good when AMD is a budget platform
>>102954546>gamingget a job incel AMD user
>>102954572It's an Intel Corporation investors problem.
>>102954639Intel CPUs are manufactured by TSMC, brainlet.
>>102954572hyperthreading loss13900k come with 14900k estrogen cores too
>>102954647Who is buying unlocked processors again?
>>102954714Since when? NTA
>>102954898Since Intel couldn't figure out anything past 14nm++++++++++++++++++ and 10nm being a massive failure.
>>102954546Had to make CPUs that didn't kill themselves, thus the lack of performance. Not the first time Intel cut corners. In the past their CPUs had huge security vulnerabilities, fixing them caused a major loss in performance. I think this time they don't only use consumers but the enterprise market as well.
>go super safe on stock clocks and voltages after the past two generations run into issues with that>also kill hyperthreading because ?????>performance isn't as amazing now>but it's a K card and can be overclocked anyway so who cares
>>102954546more like is not oc to death at the factoryi hate the retarded marketing of numbers should always go up no matter what
>>102955131>>102955182>>102955249So apart from not frying itself what does this CPU do? Why buy it?
>>102954546paper launch confirmed. cudimm memory not available beyond jedec standard junk with cas of 52
uh oh...which one do we watch first
>>102955265dunno man, its probably better at daily use, remember that the benchmarks are very biased to the peak performancealso would use less power, thus lower electricity bills, you need a smaller power supply and less heat problems
>>102954546>>102955182It's like a 3% perf drop for 50 less wattsI'm buying it
>>102955517The CPU uses wattage from the 24pin connector now.around 50w according to Gamers Nexus.
>>102955517Why not buy AMD which comes with an even greater power draw drop while performance increases?
Intel Defense Force (IDF) is already here.
>>102955517>>102955567
>>102955265>why buy itTo upgrade from a 9th gen or older cpu? It's the same way something shitty like a 4050 or 4060 are still good cards IF the purchase use case is upgrading from a low-midrange card from multiple generations ago, but are bad cards if trying to upgrade from just one or two gens prior.
implessive. is intel going to cuck amd out of the low end GPU space soon?
>>102955607but why wouldn't you just get an AMD chip instead?
intoddlers btfo
>>102954813people with jobs
>>102955650Possibly, but I don't think anyone buys anything but Nvidia
>>102954898Since right now. I think this is there first TSMC made desktop chip. Before that they had partially TSMC-made Meteor Lake on mobile and TSMC-made Lunar Lake
>>102954546
>Intel leap frogs AMD, ends up on 3nm before them>Performs worseI believed people when they said Intel was hobbled by their obsolete manufacturing process. But now there's no excuse. Their designs are just bad??
>>102955583>>102955542bruh nevermind
>>102954572Cuck cores are a terrible stop gap, but no that's not the issue here. Actually making their chips run within safe margins with any power efficiency again means regressing stock performance back to Alder Lake. Also according to leaks Pat scrapped the upcoming big architecture developments insisting the future was gpu and ai, they can keep making incremental gains each gen on cpu. Intel toxic positivity knows no limits. They've learned nothing.
>>102955898>Their designs are just bad??Bad, primarily because they decided to get into the moar coar game only way too late. The FUD they pushed for a good decade re: moar coars and DX12/Mantle held shit back longer than it needed to, which everyone ate-up gleefully because tech journos (and the rabidly pro-Intel /v/ and /g/ for local comparisons) had no problem pushing the whole>multi-core performance is useless in DX11 - the only API anyone uses - lmao what are you gonna tell me Mantle is gonna go anywhere when only EA trash uses it? And Star Swarm is just a benchmark that cheats with multithreading and a newer API, no one's dumping 7 for that, how many DX12 games are there anyway?thing for the longest time, although Bulldozer-era AMD and launch Windows 10 certainly made it easy to do so with zero pushbackThen after Doom showcased that Mantle was the beta test for Vulkan, and AMD showcased that Bulldozer was the beta test for Ryzen, Intel suddenly has to design CPU's where single-threaded performance isn't the end-all performance metric now that the plug is forcefully-pulled on Windows 7 around the same time, whereas AMD spent the post-Phenom years seeing what type of moar coar setup actually works, and when the first attempt didn't, they hire the guy who helped them curbstomp Intel for a decade to make it more like Opteron so it doesn't suck OFF paperIt wasn't just vidya either, not-Chrome browsers also painted themselves into a corner by assuming multithreading was something only extreme cases called for, ignoring the fact that the internet itself was turning into layers upon layers of CDN's overwhelming even Haswell CPU's. Firefox was horrible during that period up until Mozilla nuked the old codebase for multiprocess browsing when E10S development was as alive as ShumwayLong story short, Intel never saw a scenario where they'd have to stop cruising off of single-threaded shit, and Ryzen not sucking has resulted in panic-designing (cuck cores)
>>102954572It could be a peak core clock problem, the new P-cores can't get quite as high as the previous gen ones>>102954800I'm pretty sure it's not that, games can't really saturate 24 threads
Here we go again...Enjoy those ATX 12V power spikes Intel sissies.
>>102955388So far, JayzTwoCents has been the most realistic review.AMD uboxed and Grouchy Nexus are going to shit on anything they review made by Intel. Linus Cucktips is going to review based on what he thinks will generate clicks and likes. I think some of these reviewers lose focus that intel's main focus is to sell CPUs to buisnesses and governments for productivity. The "gamer" base doesn't really mean much to them.
>>102956223And these reviews are for how much the average person should care about them not businesses. If intel doesn't care then neither should gamers and the review will reflect that.
fake news
>>102955650intels gpu driver are worse than amds, nobody sane is going to use it
>>102955517>It's like a 3% perf drop for 50 less wattsAMD is still more efficient at both games and productivity, Intel dropped the towel hard on this one.
>>102955388KitGuru's
>force straight white males to attend to seminars where they got shamed for colonialism among many other shit>invest massively in IsraelWhat did they mean by this?
>>102956223It shouldn't take step backs in gaming even. That tells a lot about the technology overall. Half-baked.
>>102956223So they're fine losing more marketshare?
>Commiefornia and other libtard states ban high energy consuming "assault computers">Intel and AMD come out with low capacity, energy, CPUs that suck>Intel and AMD catch all the flak for suckingYou have to love them getting the blame instead of the root cause.
>>102954546Wait, i thought this was going to be Intel's comeback against AMD's anemic 9xxx launch and nuke them from orbit
>>102954546at what voltage and what 1% lows?
>>102954647Intcels were bragging for years about their powerful GAMING CPUs
>>102956777This, I still remember the StarCraft 2 720p and Skyrim 800x600 benchmarks.
>>1029567539000 came out just in time for a good efficiency boost to make Intel's new offering irrelevant.
>>102954546I bet AyyMD knew intelaviv's gen was shit so they nerfed 9000 and then linked up.Pat owes Mommy Su a solid.
>>102956823Always the same 5 Nigguherz! argument in Far Cry 5. They literally claimed the i3-9350k was better than anything AMD ever produced.
>>102956777>>102956823>>102956860That was slop journo horseshit that anyone with a brain knew was nonsenseThe fucked up thing is you will have more people now defending that type of "benchmarking" not just on intel but AMD now too.It set the idea of "gayming" CPU into an impressionable audience and now you have every other retard thinking a 14900k or X3D is going to boost their gaming 20% while on their middling 4070.
>>102955388>linus shill tipswhy the fuck would you even consider that garbage?
>>102956223Yeah, GamersNexus should focus on the value proposition of a $700 CPU for shuffling excel tables around.
Core Ultra -9%
>>102957185Clearly you only need the 5 and 7 series for gaming, just like you don't need the R9 for gaming.
>>102956732>Commiefornia and other libtard states ban high energy consuming "assault computers"That's not a thing, curb your persecution complex, Pat
>>102954572Wait for a Windows fix.
saars the reviews are increasing
Should've borrowed the glue for AMD.
>>102956223>intel's main focus is to sell CPUs to businesses and governments for productivityThe Hardware Unboxed video goes though some generic productivity workloads in addition to the gaming benchmarks.Spoilers: It's shit in every context.Like these Photoshop scores are abnormally ass, but even ignoring that the Ultra 9 285K loses to/ties the 9950X (sometimes even the 9900X) in pretty much every occasion. If I was going to buy Intel I'd suffer the dead platform and go with a cheaper 13th or 14th gen solution, assuming those stopped killing themselves.
>>102957435>13th or 14th gen solution, assuming those stopped killing themselvesWe cannot know if they have or have not until a good chunk of time passes and somebody actually follows up on doing some analysis on failure data in the mean-time. Even then it's going to be questionable because not all of them are going to get the microcode update and as such the flaws will continue to manifest. Safest thing to do is to just not buy anything 13th or 14th gen, hell I wouldn't even trust the new crap until it stands the test of time. If they fucked up once, they can fuck up twice.
gluebros...
>>102956223Nigga
>>102956223Productivity is also shit.
>>102957264>Wait for a Windows fix.Sounds likely.
>>102956223>jayztwocents>grouchy>cucktipsFrom Micro$haft to this, why do you nerd virgins always use these gay nicknames? Do you not feel embarrassed being so lame? So pathetic.
>>102954546Intel is sending me a full refund for my 14900K that got cooked. Should I>Get another 14900k chip, plug and play into dormant PC>Sell the Z790 board and go Z890 + 285K>Just get AMD please! Because gaming! lol!I have an H150i Cooler and it struggled to keep the 14900k cool. I dunno when the CPU failed to make it untenable though. I don’t game beyond VRchat so AMD’s main selling point isn’t as compelling to me.
>>102958039switch to amd, get a non x3d cpu if you dont game much.
>>102958039Get a 13600k and spend the rest on something else.
>>102958039>AMD’s main selling point isn’t as compelling to meBeing the best at both performance and efficiency in productivity and vidya isn't compelling to you? Are you perhaps mentally challenged?
>>102958039>Just get AMD please! Because gaming! lol!AMD wins or at worst ties productivity too.
>>102958039>Sell the Z790 board and go Z890 + 285K>unironically considering Arrow LakeNice bait, fell for it.
>>102954546And what of good userbenchmark?
>>102957882jayztwocents is the guy's actual channel name tho
>>102958039>I don’t game beyond VRchat so AMD’s main selling point isn’t as compelling to me.Isn't VRChat one of those games which gets insane benefits from X3D chips?lol
Tell me guys that at least intel got its shit together and avx512 is back
>>102956223>So far, JayzTwoCents has been the most realistic review.
>>102956223Their audience isn't the government, so why would they give a flying fuck about Intel trying to hawk their inferior shit to the government?
>>102958262alder lake (12000 series) has avx512
>>102958262>and avx512 is backNo
>>102958262no intelaviv is full of retards
>>102954546Apparently - in gaming, my 5700x3d trades blows with this new thing. For extra spice - I have an Intel GPU.
>>102958039outside intel i3s and n series aymd is better at pretty much everything
>>102958262What's so special about avx512? I keep seeing it mentioned
>>102955696Because AMD is shit, no matter what its fantards say.t. 2/3 of owned AMD processors cooked themselves to death on stock clocks despite having proper tower and later water cooling
>>102958571512 bit floating point registers. Its in the name. Allows some heavy optimization for certain workloads - like video encoding, encryption or scientific calculations.At the minimum - allows to fuse two AVX2 ops into one.
>>102958039VRchat is where Ryzens with Vcache roflstomp Intelt. 7800X3D VRchat player
>>102958571avx512 includes 512 big registers but no one cares about that there are bunch of new ops which also work for 128/256 bit vectors. Intel in their infinite wisdom refuse to make a half-throughput avx512 on their e cores because it would add 0.1mm to each core
>>102958738Haven't played VRChat in forever
Shalom Intelsirs, we keep winnin lol
How the fuck did they lose to a bad AMD generation?Isn't this chip the last product from the previous leadership before the CEO changed? If I recall they still had to release the hardware they developed before tearing it all down
>>102958692>>102958776Would any of you start buying Intel CPUs if they came with AVX512?
>>102958837I have no use case for AVX512.
>>102955388Whatever you little zoomie
>>102958821They both did the same thing: make it more efficient. The difference is, AMD didn't regress in any way.
>>102954572pretty sure all core ultra's cores are cucked. no hyperthreading whatsoever. it is so over for intel aviv.
>for gayming
>>102956223>Grouchy Nexus are going to shit on anything they review made by IntelGamers Nexus backs up all their claims with raw data. If they're shitting on intel's offering it means intel's offering is shit.
>>102958918Why are both brands going in that direction anyways?I know intel has a reason due to it's housefires but is something going on?Has a ceiling been reached?
>>102959090>least obvious steve fanboy
>>102958279only eartly chips and you need to do some bullshit for it work>>102958837it would be usefull to me, and i already know what is the experience of lacking an important instruction years later, i like my desktops to last very long, so yeah i would really consider it, that is why i asked in the first place if this gen has itt. phenom 1055t owner that cannot be used bc of its old instructions>>102959211we were getting to retardation levels of power consumption, and thuse heat disipation, it was starting to be a technical problem, also servers, which are the bigger market, want the efficiency and they share the inner designs
>>102954546>Loses to 14900KOkay, but does it rust and ACK itself?
>>102959211Once upon a time everything only used a couple of cores at most so it was just a matter of building the most powerful core you possible could. Now everything runs 9001 threads, so by halving your power consumption, you double the number of cores you can put on a chip and thereby double your potential performance. That's sort of how the cuck cores work. They're actually less efficient than the real cores, but because they're about 40% of the performance of a real core but only take up 25% of the space, you can fit moar coars on the same chip. That low efficiency was hurting Intel though, because more heat means they'll throttle (or at least boost less) without a serious cooling system.
>>102954640>>cpus and boards get price dropsuh
>>102954546damn, i thought the AMD last gen was bad enough with a 0% improvement in performance, but here comes intel and outdoes AMD with a negative performance improvement over the last gen.
>>102955583What's 5700x3d relative to 5800x3d?
>>102958685>Intel-fag FUDing this hard Modern CPUs don't cook themselves unless you are overclocking and overvolting them like a moron. They throttle back to safety at stock.You were arm-chairing overclocking and got burn from it like a novice who didn't understand the risks even or what they are doing.
>>102954546Only a retard would purchase an i9 for gaming.
>>102960047It is because both aren't doing crazy factory overclocking out of the box. It had CPU releases used to be before Ryzen+/Kaby Lake started the crazy factory overclocking at the expense of thermals and power efficiency. Unironically, arm-chair overclocking might become a thing again for people who want performance but care little for thermals and power efficiency.
>>102958571
>>102958571>>102958837>>102958879this is an old image, now there's more stuff taking advantage of it
>>102960766>FUDOh sure, I was just imagining the second of the two 3800xes downclocking to 500Mhz while reporting running at 115C and shutting off the computer while a water cooling loop was active and running. Shove some AMD merch up your ass, fagtard. No overclocking was involved unless you're an ape and think default DOCP allowing DDR4 to run at 3200 instead of 2133 is "overclocking". The CPU parameters were untouched.
>>102961448>brainlet likely left plastic cover on their water-block on or didn't mount it correctly100% skill issue, you can't fix stupid or carelessness
>>102958837If the platform costs were cheaper than am5 yes.
>>10295992112900K AVX-512 chad hereYou need an msi z690 board, an cpu manufactured in 2021 and uuh you change some BIOS settings, maybe flash the BIOS to an earlier version first depending on the board, and that's about it. You disable all the e-cores, save exit, go back in and enable AVX-512. Happy RPCS3ing.
>>102956223retard
>>102961448I was on a 3600 that never went above 70C on a 212 for years. I have no idea how you got multiple AMD cpus to blow up on you, but it's probably PEBCAK.
>>102961651>>102961699>least obvious amdronesPathetic.>plastic obviously off>thermal fucking grizzly thermal paste>solid connection >fan curves set faster than should be necessary>issue occurred on two separate chips with both the provided wraith cooler and a high end water cooler>amdrones claim nothing happened and if it did, ti was user error
>>102961671>AVX-512doesnt nue intel have AVX10.. which can process AVX-512 style work
>>102961727Did AMD warranty it?If yes, who cares. Shit happens. If not, why not.>thermal grizzlyWhy. And which one.More importantly, did both issues occur on the same motherboard? Did you consider changing the motherboard's voltage settings at all or perhaps buying a new one after the first occurrence? I can tell you I've seen some motherboard overvolt Zen 2 to an extreme degree.>>102961753This is Intel's official Ark website. Do you see AVX10 support in here, anon? Because I don't.Also, AVX10 is a SUBSET of full AVX-512. However you slice it, it's a downgrade.https://ark.intel.com/content/www/us/en/ark/products/241060/intel-core-ultra-9-processor-285k-36m-cache-up-to-5-70-ghz.html
>>102961727Not AMD-drone you brainlet, it is clearly skill issue by the end user. You improperly mounted the water block. Didn't matter what CPU you would use, it will have throttled like hell.
>>102954546barely lake 2014 - 2016doom lake 2017 - 2022nuke lake 2023 - 2024no longer lake 2025
>>102954640it's the same msrp as a 14900k you retard
>>102954546So THIS is the power of TSMC...
>>102961929>prices can never change from MSRP
>>102954572They pussied out and tried to go for efficiency because europoors were scared of the power. I love my 14th gen and seeing it pull 320 american watts out of the wall makes me happy.What exactly is the use case in buying the 15th gen intel? 14th gen will be on sale and be better performance. How retarded are consumers.
>>102955265Compared to any other cpu in the market it’s 5% faster at compression and other random “productivity” tasks that nobody else does.
>>102959090>trust the scienceMan you fags are pathetic. Fact is an intel cpu is just better in real life, I’m not poor nor am I brand loyal, I have built a several PCs in the last couple years and can tell the difference.
>>102954546can someone explain to me why intel hasn't done what AMD did decades ago by selling its fab division as another company (global foundries)?
>>102962101Intel has latency issues to solve, just like AMD did when they first introduced Zen.
>>102962426this was made by tSMC
>>102955265It's for people like this who can't help it >>102962231
>>102955265for people who use their builds for work and internet browsing only. if you don't play games, its very good
>>102962426Doesn't make as much sense for intel as other than CPU/GPU they also make a wide range of other things like NICs, Wifi cards, and even things like the now defunct Optane and many other things. Also would you really want even more of the chips to be relying on TSMC which is mostly situated in Taiwan and can be gone at any moment if CN were to attack. With North Korea and Russia in a treaty with CN, South Korea Samsung fabs would also be gone, leaving the world with only Intel left as the sole maker of high end chips, it's a national security issue.
>>102962117>muh price cuts on botched CPUsyou do realize how unbelievably retarded you sound, yeah fag? and there was barely any drop at all in prices, even during the entire controversy and scumfuckery, right up until they launched this super shit lineupstop gobbling cock
>>102954572Its a p-core problem. tsmc-3 was a new process to intel and they didn't really know how to squeeze the most out of it, that probably left a few hundred MHz on the table. Also the chiplet architecture; latency to memory is up.But down the product stack, the chip looks great. I assume this is going to sell pretty well. Its is going to be cool and quiet and high performance and have the built-in NPU and reasonable integrated graphics.
>>102954640kek some next level cope
I think it is odd they stopped the model # at 285. To me that suggests they deliberately left room at the top of the stack. I doubt intel would do it, but what if they take their brand new 18A, and leverage their chiplet manufacturing, and just stick a new compute die down as a low volume halo product. That would be a chad move; it would show confidence in their 18A, show the utility of their chiplet design techniques, grab mindshare, etc. Gamers are a small market, so for a cash strapped company, this might not make sense, but Pat, if you are listening, I want this.
>>102962760Maybe the ks will now be 290 in the lineup.
>>102962486then sell SOME of the factories?>they also make a wide range of other things like NICs, Wifi cardsAFAIU they have (had?) factories in china for that some of shit too. hell they made the Edison modules in china IIRC.
>>102955517>performance drop generation-to-generation>I'm buyinggrim
>>102954714Like those imbecile idiots in government even know or give a shit.
AMD has shit chipset/boards , maybe this will change when the B850 boards come out but as of now. arrow lake has better boards /IO
>>102962231https://youtu.be/uc6f_2nPSX8?t=187
>>102963121Doubt it, the X870 boards are just the B650E boards rebranded with mandatory USB4 and jacking the prices way up, the fucking X670 and X670E boards were rebranded into X870E
>>102962129>europoors were scared of the power.>pull 320 american watts out of the wall you guys literally get 120v instead of 240v though.. lol
>>102962101DEI designed that cpu.TSMC just made it.
>>102960840don't deadname Ultranny 9 like that
>>102958685>make shit up>post itMany such cases
>>102960766...or your board has some electrical problem and ignores what CPU demands, voltage wise.
>>102961727Mate. There are literally tens of millions Ryzens running right now without issues. Even in the times of original Athlons, where CPUs would happily cook themselves, it happening was rare as fuck. What you are suggesting is so out of ordinary that it _has to_ have been some odd, isolated, incident.
>>102962101x86 is doa
>>102962594only one tile of the cpu is 3nm the other are 6nm
>>102954546when will people learn? Intel made one good CPU series in their lifetime, AMD made one bad l.
>>102963170>X670 and X670E boards were rebranded into X870EX670 has no direct 8xx counterpart, X870E is just X670E rebranded and makes USB 4.0 and Wi-Fi 7 support mandatory. It is the only dual PCH platform for 8xx chip series.
>>102960766That's too much credit, he's just lying
>>102962426These chips are 100% tsmc, what makes you think Intel can even survive as a fabless company
>>102962486>leaving the world with only Intel left as the sole maker of high end chips, it's a national security issueIt would be great for muttland because it would mean that the entire west would be forced to pay extortionate prices for chips, because supply and demand, goy. Jews rubbing their hands as we speak. WW3 in 2 more weeks.
>>102954572Nah it's an intel chiplet growing pains problem, and them thinking everything needs a NPU in it.
>>102955650except radeon 890m for example is somewhat competitive with rtx 3050
>>102961448>3800xes downclocking to 500Mhz while reporting running at 115C and shutting off the computer while a water cooling loop was active and runningMy 3700x happily ran with a fuma 2 and 3700MHz until I put it in my NAS after getting a 5900x. The 3700x still only gets to 80C or so under full load with a low profile cooler and the 5900x also never exceeds 85C at full load. Learn how to mount a cooler or don't use a shitty board that overclocks without asking. A chip can only do so much against PEBKAC.
>>102966540>3700MHz3700MHZ DDR4
>>102961448Docp is overclocking periodThere are ram kits that work at 3200mhz ootb without needing xmp and yours isn't one of those
>>102954898since 2009 at least: https://www.forbes.com/2009/03/02/intel-tsmc-atom-technology_atom.html
Why is this board obsessed with defending Intel no matter how bad things get? Is it just another brand of the contrarianism you see on /v/?
> have to buy a new motherboard because of another socket change> have to buy new DDR5 memoryGuess you can reuse -- your power supply?
>>102954546omg a non-gaming CPU full of neural cores is worse at certain gaming tasks CALL THE PRESS!synthetic benchmarks are dumb.
>>102962426Now? Intel is on the hook for billions in Chips Act graft. Don't think spinning or selling their fabs is an option:https://www.intel.com/content/www/us/en/newsroom/news/us-chips-act-intel-direct-funding.html
>buy best of the best>upgrade once every 8 yearswhy would you pretend like a mobo supporting multiple cpu gens even matters
>>102966891>neural coresThat sounds profoundly useless. Why is it there?
>>102967069>neural cores>Why is it there?Because in very near future every CPU is going to have a weird combination of matrix math, analog compute, and neural acceleration cores.You can get m.2 key neural accelerators for $25. You have to talk to a sales rep and get deep into it to get a quote from Mythic on their analog compute/matrix cores...
>>102967027Because on AM platforms it has been a really good deal in the long run. You don't need to go full retard and buy expensive and stay on it for so long.
>>102967217>You don't need to go full retard and buy expensive and stay on it for so long.It's cheaper that way? I just upgraded from i5-6600K to i9-14900K, and prior to that it was an i7-930... I paid a measly $450 for this new i9 oh noes
>>102967315yeah amdrones are retarded anon that's not newi personally went from a 4790k to a 12700k, at this rate i will never need to upgrade again
>>102967315You went from a 4 core with hyperthreading to a 4 core without hyperthreading. You are clearly an idiot.
>>102967336zoom zoom detectedHyperthreading didn't mean dick during the time
>>102967315you bought a 14900k. that automatically voids the validity of your opinions
>>102967336My use case scenarios change. What is your use case scenario? How many AMD generations do you skip? I've purchased 3 whole CPUs in the last ~15 years how about you?>>102967361I waited until last week and the motherboard's firmware was patched to Intel specifications. Even though I have a K series I'm not even using XMP. I'm running entirely within Intel specs. Sucks to be you.
amdrones needing to larp as imaginary intel users sure are funny
>>102967359Anon, "that time" was 2016. You paid damn near 300 bucks for a quadcore without hyperthreading the year before Ryzen came out. You're retarded. I was there. Everyone in 2016 knew quadcores were on the way out. Everyone with a brain, that is.>>102967377I bought a 1500X to get me started in 2017 when I got my first job, because the 1600 was out of stock everywhere. I got a 3600 as a same socket upgrade in 2019 and it served me well for 5 long years. My AVX-512 enabled i9 12900k is in the mail, and the only reason I'm really getting it is to run RPCS3 on it. Will also serve me better when I upgrade my GPU hopefully on Black Friday, if not then maybe next year.
>>102967409>he bought a housefire starterLMAO ENJOY THROTTLING EVEN WITH AN AIO
intel APO saar
that anon is arguing with himself btw, actual mental illness
>>102967416Thermalright Phantom Spirit. Even under prime95 it won't throttle. I've seen the benchmarks. And I plan to turn off the cuck cores so it's not going to throttle under any situation whatsoever.
>>102967409>without hyperthreadingI was overclocking the shit out of the i5 and mostly doing gaming/internet stuff. You must not game because you keep crying about core/fakecore counts...Everyone with a brain knows you don't waste money trying to overclock unused cores.
>>102967430>phantom spirit and prime95 in the same sentenceit's not even funny i just feel bad for you lil nigga
>>102958052even if you don't game, 3D cpus are very useful. more chache means you can do simulations etc. much faster because you can cache more data into the CPU, process them in parallel etc.
Hailo-8 M.2 AI modules max out at only 26TOPS already being made irrelevant by Intel onboard lol.The "Hailo-8 Century High Performance PCIe Card" tops out at 208TOPS but that is also irrelevant compared to modern Nvidia offerings.
Bought a brand new 12700k for $125 recently and it outperforms Intel's entire current generation kek.
>>102967446You might just be the most retarded beast on God's green Earth.
>>102967485what exactly is showing me some rando benchmark of stock 6600K supposed to mean? that I was wrong to buy it and overclock it and use it until last week?kys
>>102967482what's the sp score on that badboy?i got one at launch with an sp score of 96. ocs to 5ghz p 4ghz e at lower than stock voltage and has been stable for 2 yearsalso got some dank ddr4 4000 ram so i am set for the forseeable future
>>102967474single core performance saar
>>102967315>I just upgraded from i5-6600KFor years before you upgraded you were on a slow obsolete CPU, and we both know it if you did anything other than sit there staring at a web browser.Anyway, mid tier CPU upgrade to another mid tier is way better than high end and then wait twice as long. The amount of processing power available in the spectrum of one generation of CPUs is always lower than one generation compared to subsequent generations, even if only comparing different mid tier CPUs.
>>102954546Could one say that Intel kinda BULLDOZED this one?
>>102967482>12700k is faster than intel's current fastest cpuhow did they go backwards
>>102967551Keep trying to make me feel bad about the past it won't work retard. I had years of cheap fun overclocking so you should cry about the 8 AMD cpus you bought in the same amount of time I bought 2 Intel cpus keki5-6600K was great for overclocking and no matter how hard you cry that historical fact won't change.
>>102967510Nothing can save you from your retardation. Some retarded niggers like you are simply beyond saving. You're a nigger, and you will be a nigger until the day you die. Kys.>>102967551Don't bother, the man's a nigger.
>>102967584Overclocking gained the illustrious 6700K an entire 6 fps in the benchmark I showed. Turning on hyperthreading on the same architecture gained 30. You're a retarded nigger.
>>102967584Not as good as Bulldozer at overclocking. Who cares. Your CPU was not good at anything for years.
>>102955388watch the Fuster Cluck one from Kitguru first
>>102954546They used up all their silicones into making chips for guided missiles.
>>102967551>mid tier CPU upgrade to another mid tier is way better than high end and then wait twice as long.This totally depends on the progression of CPUs (It is far from linear) and how long you're waiting. And platform costs, if you need a new mobo, if you need new ram and if the ram available at that time will be good for the lifetime of the DDR generation.Upgrading just one generation is very rarely worth it though.>Upgrading from 2-5600k to 6600k -> pants on head>Upgrading from 6600k to 7600k -> pants on head>Upgrading from 7600k to 8/9600k / 2700x -> GoodAdvanced examples, CPUs are both 2 gens ahead and have much better IPC and clocks but mobo & ram situation is different:>Upgrading from 1700x to 3700x meh cause you have pretty bad ram and mobo.>Upgrading from 2700x to 5700x Good because you have a good mobo and good ram.
>>102966891The NPU is maybe 3% of the die space and is separate from the actual cores, stop coping
>>102967713Both AM4 and AM5 allow for one amazing upgrade. Like you said Zen+ for example to Zen 3 (X3D), or Zen 4 to Zen 6. It's just unbeatable by any Intel waiting game.
>>102967595>>102967617>>102967657Stay mad at your past 4 annual cpu purchases AMDtards.
>>102967791Go back to losercopemark please.
>>102967584>>102967617>i5-6600K was great for overclockingIt's a CPU for retards. The only reason it "overclocks well" it because intel cared more about TDP than performance cause AMD had not competed for many years.The intel quadcores were a good test to see if you had any clue about the CPU market at all.>2000 - Smart consumer>3000 - late upgrader, still smart>4000 - possibly still smart, but not by much as you could get full 2500k/2600k systems for less than a 4690k + mobo ATP>5000 - somewhat clueless FPS chaser>6000 - clueless consooomer>7000 - Even more clueless consoomer.
>>102967807You posted a benchmark showing a 6700K @ 4.8ghz and a stock 6600K (3.5ghz). You're dumb. I said I was overclocking, no one that says overclocking means less than 50%. You can shut up and quit crying about my old CPU and start worrying about which 3 AMDs you're buying for 2025, and then the 3 for 2026, etc, etc, etc.
>>102967827I already said I own a 12900k, now go back to niggercopemark where you belong.>>102967825this.
I’m running on 2600k and can’t decide what to go to. What is today’s 2600k?
>>102967838>I already said I own a 12900k, now go back to niggercopemark where you belong.You're a retarded liar that can't keep track of your own lying ass lies. You said you ordered a 12900K. I own a 14900K now you can stfu and stop crying about the 6600K anytime you want faggot.
>>102967860Yeah, I ordered and paid for it. That means I own it. Faggot. It's mine. I paid for it.Now stop with the semantic bullshit and go back to niggercopemark.
>>1029678545800x3d unironically beats the 7800x3d in some stuff and probably the 9800x3d toosuffice to say amd will not be making that mistake again in the future
>>102967869>it's in the mailYeah sure. Keep crying about my 14900K and post more completely irrelevant benchmarks AMDtard kek
>>102967827>6700K @ 4.8ghz and a stock 6600KHe posted a bench with a 6700k OC (4.8) and stock (4.2). It showed the increase was small.>(3.5ghz)6600k boost freq is 3.9 and that's what it ran at unless you had shit cooling.>I said I was overclocking, no one that says overclocking means less than 50%hahah, what? No one that says OC and actually uses the thing these days means near 50%.Are you saying you ran your 6600k at 5.2 or 5.85Ghz? It doesn't run at that speed without LN2. Don't make a fool out of yourself, anon.
>>102967788>Both AM4 and AM5 allow for one amazing upgrade.Eh, maybe. zen1 was great at the time but didn't allow for a good upgrade without new mobo & ram.And AM5 isnt super worth buying compared to am4 for midrange still.
where is the single core performance comparison (only thing that matters)?
>>102967980still loses to like a 12600k lol it's actually bad not a meme
>>1029679497700X to 11800X3D sounds like a wicked upgrade to me.
>>102963278yet we dont' have retarded power laws like you. Also you're poor
>>102967980what's your chosen single core benchmarkit can be on par with 9950X in SPECint 1T depending on the compiler flags
>>102954800hyperthreading is not important for gaymen, if anything you want to disable it because of the introduced latency. Despite the HT loss, it matches or beats the 9950X in a bunch of MT workloads like compiling. The issue seems to be latency and/or scheduling related due to the new layout. I would like to see benchmarks with e-cores disabled, maybe we'll see software fixes in the next few weeks if possible.
>>102966918no one would buy anywayhttps://www.tomshardware.com/tech-industry/tsmc-is-not-interested-in-buying-intels-fabs
>>102967980>(only thing that matters)lel, no
>>102961798Hydronaut, then later Kryonaut when repasting was attempted. Piece of shit cooked the paste, both times. Why? Because it's supposed to be good high end paste. What the fuck do you mean, "why"? The mobo was replaced for both cpus. Same issue. Even disabling DOCP and running full stock (I love buying ram and being forced to run it at a lower speed) did nothing. Total failure occurred only after warranty expiration, but prior to that both chips ran hotter than anything ever used before or after.>>102961882No, I did not. The same issue happened with the stock wraith cooler on the first card. Gonna claim that was "improperly mounted" too because you need to defend your corporation? Meanwhile I have an intel system with a noctua cooler that had its fan mounted backwards for the longest time because it was built under poor circumstances, and with it backwards that 6700k never even hit 62C under heavy, full-core load. Now the 13th gen i7 I have has never gone over 80c for more than a split second even with an inadequate single tower noctua. Something doesn't add up.
>>102967825>you HAVE to buy 2nd gen and stick with it forever>noooo you can't skip it and get a newer variantWhy do you povertyfags always pull this shit? Nobody cares about your awful 2500k and how much you think it's hot shit because it was only a little outdated when mommy bought it for you.
>>102968466those pastes are known for drying out in 2 monthsyou're just stupid
>>102968538>anon buys something high end but advertised as a general-purpose product>expects it to last a good chunk of time, at least as much as a lower end product>it's anon's fault when it doesn't last and not the fault of thermal grizzly for very blatant false advertisingYou niggers will do anything but blame a product
>>102954546Noob here, can we give it a benefit of a doubt that this was caused by very early release with shitty compatibility with current games and apps, or is the chips already fucked and there's no saving it, and we're going AMD this generation?
>>102968712the latter
>>102955388For me it’s the autistic German, he did a pretty good job of summing it all up and pointing out the benefits. Either him or Wendell, with the latter being more of speaking about the CPU rather than flooding your screen with numbers.>JayzTwoCentsMidwit take for the uninspired masses
>>102968538Known where, your ass? That shit isn't detailed anywhere, and isn't called out in any storefront reviews.
>>102968684he blamed the product (paste). but other anon is blaming another product. both these niggas blaming products. what are you smoking
>>102968269Noita
Surely the terrible gaming performance is a software issue and can be resolved with microcode updates, rather than being a design flaw with cache latency
>>102962594You don't GET ITIf it's not THE BEST then it's SHIT.HOW MANY times do we have to TELL YOU
>>102955388It's hilarious, but at this point I trust LTT more because AMDUnboxed sucks and GamersNexus is just trying so hard to become a drama channel, at least LTT is too incompetent to have an agenda. Note that this wasn't always the case, GamersNexus and Hardware Unboxed went crazy after Nvidia stopped giving them free GPUs
>>102969250>GamersNexus is just trying so hard to become a drama channelLevel with me for a secondIs it drama because you believe all of the reporting is clickbait as you are you are just not aware of just how bad the business practices of some companies are? Or do you already understand how bad things can be and have internalized and thus excused it already?As someone who has been keenly aware of just how garbage some of these companies are it's been a hell of a thing to have someone actually report on it.
>>102962594>tsmc-3I love you for this
>>102954572is there even a good way to programmatically query which core is which, for setting affinities, or are we now at the mercy of the schedulers?
>>102969250>I trust LTTyou trust corporateshilLTechTips? why?
>>102962594>the built-in NPUNobody cares about it. No serious software even uses it. It's there because Microsoft wanted it for the whole "AI PC" marketing bullshit.
>>102954546>Core Ultra 9 285KWho the fuck is coming up with these names? What the fuck is Ultra now? Why do the numbers end with 5s instead of zeroes?Fuck this gay ass jewish company, it's dead anyway
>>102968211Of course, 7700x to anything after 2017 is a great upgrade but that's because buying it was dumb.>>102968520>you povertyfagsPeople with money agree and never bought the shitty quadcores, they bought HEDT. Or you know, they just agree that they threw away $300 cause they don't care. You're just mad about being called out for a really stupid purchase decision.
>>102969448>illusion of binary choiceGamersNexus wasn't supposed to be a drama alert / bad business practices channel. It was supposed to provide accurate, dependable reviews on products. Their latest shift towards drama is, at best an attempt at clickbait, and at worse an attempt at sensationalism. I'm not mad that they report on it, but my trust towards their latest content has diminished greatly, I can't with a straight face believe their claims at face value.
>>102969492They're too retarded to have any agenda. Their style of videomaking lends itself to retarded videos. I think they're more likely to fail into being right than succeed at being wrong
>>102969613>>102969631It's (mostly) not drama they're just covering companies providing very bad products. Actually, they copied coffeezilla a bit. They're not super relevant products though, we know AIOs suck, we know prebuilts suck, it'd be a lot better and interesting if they covered how nvidia/intel/amd etc fucked us since it can't be avoided. But they're not lying or making stuff up.They're pretty meh videos though.
>>102969565>um ackshually people with money and clout and intelligence and class agree with my retarded assessment
>>102960115Slightly below it; there's virtually no difference between them.
>>102969613>Their latest shift towards dramaTheir LTT video(s) are the only instance that qualify in recent memory as something close to drama, and that appears to have been largely a one-off. EKG reporting isn't drama, nor are other similar posts for Asus et al. >It was supposed to provide accurate, dependable reviews on products.They do.I sense some confusion here as to what Gamer Nexus is. Gamer Nexus describes itself as a "PC Hardware Journalism" business, and has expressly had that label since at least 2018. Hardware reviews and benchmarks apply, but that's not the totality of "PC Hardware Journalism". This is why they cover malpractice stories and also aim for factory tours, interviews, etc.
>>102966870>400w cpu + 600 w gpulol
>>102955719inside of them?
>>102969900Topkek, yes that is what you said because you don't actually have any valid argument against what I said in the first place, because it's right.
>>102954640>cpus and boards get price dropsYou can get 5700x3d and a bargain bin AM4 board for dirt cheap and still beat Intel Cuck Ultra
>>102955284The Digital Foundry one
im not changing my 7500F anytime soon
>>102970343>it's right because I say soMake another "never obsolete" thread about the 2500k you poor retard.
>285kwhat the fuck is this shit? i dont care about intel for years but is this a new naming shit? is this the 15900k?
>>102970650No it's a 9700K with e cores
>>102970641hahah, read my post. By point it 2500k was bad (but so was all of intels i5s) until 2017 (7 years ago now) after that it's pretty shit.
>>102970982>By point it 2500k was badit wasn't you fucking imbecile. You don't need all the bells and whistles
>>102970650Seems like. It's the Core Ultra 9 flagship. Kinda strange they didn't call it 290K tho, maybe that name will instead go to the 15900KYS equivalent?>>102970933No, this is Patrick.
How do we fix Intel?
>>102954546It's just going to become a state owned concern like Boeing.
>>102958685>t. 2/3 of owned AMD processors cooked themselves to death on stock clocks despite having proper tower and later water cooling>>102961448>Oh sure, I was just imagining the second of the two 3800xes downclocking to 500Mhz while reporting running at 115CThis has been technically impossible for over 20 years.Modern processors have several 100 temperature probes build in and very simple hardwired powergating circuits which evaluate the CPU state at a very fast rate and downclock and powergate to prevent excessive temperature.Most of these can't even be manipulated since they're fused off from the factory.
>>102954546>be intel>get cucked twice by Jim Keller for the short time he worked for AMD >both times he shit a new architecture that not only completely destroy you but is so scalable they can use it with little to no modification and get over 200% IPS gain>the only downside is his design look strange at first>OK we will hire Jim Keller>he shits a design that make the kikes and streetlights scratch their heads but even an early ES give good results>kikes resume their usual backstabbing polices and the project gets canceled push p cores and cuck cores but this time they don't even have HT>launch half backed product with the same performance and power efficiency as mid 12th gen parts>it's anudda shoah
>>102954572Nah, Intel just can't keep up with AMD.I'm keen to see overclocked + tuned RAM results though.People overlooked the 9700X and 9950X because of normalfag "YouTubers". Might be the same case here.
>>102954640Enjoy your new socket in 2 years time that will have 2 more pins but you are used to this.
American fabs talk big shit till all they do is design work and suddenly realize all the actual manufacturing is in asia. May our environmental and labor laws are too stringent. Or our capitalistic overlords can't think beyond their own self interests unlike some small island nations who's very survival depends on having bleeding edge tech manufacturing. I hope asia can defend itself against china, because america ain't certainly doing it.
>>102973542Gigantic cope.
