>>101527596
>However, the microcode update will not repair impacted processors
NO REFUNDS

>>101527634
just buy a new one off amazon. send old one back to amazon for a refund.

>>101527596
>the microcode update will not repair impacted processors
Of course not. As wisdom teeth taught us, once something is impacted, the only fix is to extract it.

>>101527634
>>However, the microcode update will not repair impacted processors
Well, yeah, the higher voltages cause permanent damage.
But if they are under warranty, they should be refunded. Most countries will enforce this. Some countries with better consumer laws will force warranty replacements up to 5 years from purchase, even if the warranty was 1 year to begin with, due to the explicit defect discovered.

>>101527654
a lot faster than doing a warranty claim.
>newegg
>best buy
>amazon
>micro center
>etc
just buy a new one. package old one in its original box, and send / take it back for the refund using the new one's receipt.

>>101527654
lmao good luck making the cpu match the box serial number

>Buy another CPU, goyim. Are you poor or something?
Keep spending money on Intel Corporation defects, the cash helps fund the land stealing going on in Palestine.

>>101527697
no one records the serial number. at least here in the states.

>>101527697
see >>101527683

you don't toss the boxes right, anon?

>>101527704
>>101527714
Sellers record the serial number and even have the serial numbers they have bought, good luck trying to niggering the system, hope you like the fraud charges

>>101527726
lol no they don't. at least not here at ANY store in the states.

i've been doing this for over a decade. from fry's until they went bankrupt, microcenter, amazon, newegg, best buy, and even fucking wallmart. no one goes to those lengths here. i gave up doing warranty claims a long time ago when i realized i can just buy a new one and send the old one back for a refund. so much fucking faster.

>>101527738
Fuck off glowie, no one gonna get indicted because your "lol it totally works", go be a nigger somewhere else

>>101527746
are you mentally retarded?

>>101527749
are you a nigger?

>>101527596
Surprised they even made a statement. The smart play was to stay quiet and silently RMA.

>>101527809
they're probably going to release a statement when the new microcode drops that they will accept warranty claims for anyone suffering issues. i just hope they lax their warranty departments to not give people the second degree.
when amd had their segfault bug they made warranty claims super easy. just send them a request saying "segfault" and they shipped a new one asap to the address you gave with a return slip to send the old one back once you received the new one. i hope intel does the same.

>>101527683
i did this with my last msi motherboard. i broke its pci-express latch and got some thermal paste in the am4 socket. so i bought a new one and sent my old one back for the refund.
>>101527726
most stores here in america have a no hassle / no questions ask return policy. some might just ask if it works or not to see if they can put it back up for sale but that's it. most don't check anything. few that do only check for obvious physical damages. like seeing if you punched a hole or tossed it out a four story building. but most don't check at all. we just liquidate everything or try to resell it if it doesn't look broken.
i once bought and returned a leaf blower that i screwed up its fan blade by sucking up rocks. they don't care at all and gave me a full refund

>>101527596
>We're told that the microcode patch will not repair processors already experiencing crashes

No wonder why I can hear the hand rubbing from the class-action lawsuit lawyers.

>>101527869
Wow I didn't knew niggers browse this place

How much performance loss are we going to see for this? 5%? 10%?

File: 1719569509999801.jpg (154 KB, 1334x750)

154 KB JPG

>>101527726
for that to happen the employees have to be incentivized to give a shit retarded rich jew. you removed all that because you hate us remember?

>>101527596
>Intel's advisory says an erroneous CPU microcode is the root cause of the incessant instability issues. The microcode caused the CPU to request elevated voltage levels, resulting in the processor operating outside its safe boundaries.
I would be thrilled to know those safe boundaries, Intel. It would help to know for people can see if their CPU has been affected by this or not.

I think I made the right call only getting a 12400 and I think I'll make the right call again when I go for AMD next time.

So if this microcode is going to reduce voltage does that mean those chips will no longer be able to hit their advertised speeds?

And it's just a coincidence this microcode update is also the exact solution Computer Jesus said Intel would do if there was some sort of oxygen introduced into the manufacturing process which would cause a mass recall?

>>101527939
It took them till now to realize their processors were overvolted? Did no one at Intel bother to check their own shit? It would be very nice to know what the safe voltages are post to be. Why is it so hard for them to tell us?

>>101527596
>by mid-August
That would be another month, so let's hope the fix will arrive before then.

>>101527596
With a patch comes lower clock speeds and they all have to be reviewed again.
6ghz kek not anymore

>>101527980
>The microcode caused the CPU to request elevated voltage levels, resulting in the processor operating outside its safe boundaries.
From what it sounds like, Intel is making the claim that the firmware is causing the CPU to request more voltage than it needed. So if the CPU needed say "1.43 volts" it was requesting "1.53 volts" instead. Or something like that.

>>101527634
The article literally says that if you're having issues Intel will RMA you.

>>101527596
>For now, it is unclear if CPUs exposed to excessive voltage have suffered from invisible degradation or damage that hasn't resulted in crashes yet but could lead to errors or crashes in the future.
It depends on how much damage was done. It could be "instead of lasting ten years, it will last for 9 years and 7 months" to "in three months time, you will start having crashes that will progressive." That's the thing about degradation. No one really knows unless you can get it under a microscope and see the damage to gauge how bad it is. Personally, if you have a 13th / 14th gen, when the new microcode lands, issues a RMA / get a new one.

>>101527596

even point to point soldered tube amp users discuss things like crosstalk, intel engineers think twenty angströms and nod to eachother

>>101527596
When I was younger I expected all sorts of defects from cpus and chipsets on laptops because oems were cheap asses who didn't want to pay a few more cents to solder things properly. It was normal to hear about RMA on defective parts to replace them with defective parts, but this shouldn't be happening on desktop cpus. I avoid laptops for that reason but it looks like now I have to wait a year or two before buying hardware to be sure it isn't a dud.

>>101528004
>>101527980
read the article.

>We're told that the microcode patch currently doesn't exhibit any adverse performance impact (i.e., the chip running slower), but testing is ongoing. We can expect Intel to share more information about performance in the future.

12700K chads stay winning

>>101528070
Horse shit.
If it's voltages that are killing cpus and causing instability the only recourse is lower the voltage thus lowering clock speed.

>The company had previously advised its customers to stick with the basic power guidelines for its processors, rather than running them at fully unlocked settings, as it worked through the issues. Those instructions, which you can see here, remain in effect for now, and Intel hasn't issued any new workarounds for impacted customers. It is unclear if Intel will lift the existing restrictions after it issues the patch.
it seems like if you run the cpu within the recommended intel baseline power profiles, it appears to keep the processors, or at least, reduce the number, who go over (or stay to long in) the unsafe voltage limits.
i think buildzoid was right when he said you shouldn't be going over 1.45v's for sustain and keep it in the 1.3v range for sustained. 1.5+ voltage range really is what was frying them.

>>101528089
read the article:
>Intel's advisory says an erroneous CPU microcode is the root cause of the incessant instability issues. The microcode caused the CPU to request elevated voltage levels, resulting in the processor operating outside its safe boundaries.

>>101528089
>If it's voltages that are killing cpus and causing instability the only recourse is lower the voltage thus lowering clock speed.
No.
Voltage does not directly correlate to clock speed. You can have higher voltage than required for the clock speed you are running.

>>101527908
single core: 20-30%
all core: maybe 5%?

they are likely going to limit the maximum voltage to ~1.3 V, which is in line with 12th gen

>>101528127
>Voltage does not directly correlate to clock speed.
No?
Set your vcore to 1 volt and try for 6ghz
Report back with your findings tard.
>>101528102
Again voltage to reach 6ghz boost speed was insane.
You're gonna see lower clock speeds or more BSODs.

>>101528181
read the article: >>101528102
>>Intel's advisory says an erroneous CPU microcode is the root cause of the incessant instability issues. The microcode caused the CPU to request elevated voltage levels, resulting in the processor operating outside its safe boundaries.
>The microcode caused the CPU to request elevated voltage levels

>pump voltage to the sky to advertise high clock rate
>fry cpus
>says it will fix by lowering voltage
>it doesn't reach advertised clocks anymore
shalom

>>101528198
read the article:
https://www.tomshardware.com/pc-components/cpus/intel-finally-announces-a-solution-for-cpu-crashing-errors-claims-elevated-voltages-are-the-root-cause-fix-coming-by-mid-august
>Intel's advisory says an erroneous CPU microcode is the root cause of the incessant instability issues. The microcode caused the CPU to request elevated voltage levels, resulting in the processor operating outside its safe boundaries.

>>101527596
there's still time to buy intc. don't miss the ride to 200

>>101528205
you do it for free or are they paying you?

File: BfbZJotajpiM4woY.jpg (102 KB, 886x520)

102 KB JPG

>>101528205
Probably rushed out due to pic related, glad there's a quick fix.

>>101527869
Why do US stores operate on a high trust philosophy when the society are low trust scum baka

>>101528205
>The microcode caused the CPU to request elevated voltage levels,
It had to be able to hit advertised boost clocks so it needed all that voltage, yid.
Lawsuits incoming.

>>101528172
That renders the CPU useless. You buy this CPU for the fast single-core performance. At that point, you might as well buy an older CPU that's guaranteed not to cook itself to death over time.

>>101528235
>>101528250
do you two fail at reading comprehension?
>Intel's advisory says an erroneous CPU microcode is the root cause of the incessant instability issues. The microcode caused the CPU to request elevated voltage levels, resulting in the processor operating outside its safe boundaries.
why is this so hard for you to understand? the cpu was REQUESTING MORE than it was programmed to request.

IF ITS VID TABLE STATED DRAW 1.37V'S FOR 5.7GHZ IT WAS DRAWING IN EXCESSIVE TO 1.37V'S. enough, to put it in unsafe territory.

so instead of drawing 1.37 volts it was drawing say 1.47 volts. how is that so fucking hard to understand?

>>101528271
You'll be back here in August swearing up and down that up to 6ghz meant 6ghz for a fraction of a second.

>>101527596
So if it was just caused by this why the silence? why did they take several months to finally speak?

>>101528282
dear lord you inbred retards are something else.

>>101528260
i'm just speculating, but i don't see how there is any other solution for intel
alder lake and raptor lake is fundamentally the same silicon on the same node, for raptor lake they upped boost clocks, ring clocks and added more cache
now one of those advantages are gone (and you can overclock the ring on alder lake if you turn off the meme cores)

>>101528271
And what it was actually hitting 5.7Ghz BECAUSE it was drawing more volts than it was reporting? Did you consider that?

>>101528284
>why would any (INSERT ANYONE AND ANYTHING HERE) not want to acknowledge a failure
you're on 4chan. you see people here argue tooth and nail to not admit they are wrong about something. why would you think companies and the like be any different? they are all owned by humans after all.

>>101528300
again, read the fucking article and read about vid tables you inbred, redneck, shitstain.

>processors programmed to draw 1.37 volts for 5.7ghz
>firmware had a bug that caused it to request more than was actually programmed.
>if processor requested 1.37 volts, firmware put out a request for, SAY, 1.47 VOLTS

how can you not fucking understand this?

>>101528286
>Muh too much voltage.
You think Intel whose made cpus for 40 years "accidentally" programmed the microcode to use too much voltage to reach advertised clock speeds and that a simple adjustment will allow it to run at advertised clock speeds and at lower voltage.
Which one of use is crazy?
I'll see you again in August when you defend Intel some more.
Can't wait for the updated reviews where the 13900 and 14900 barely beat a 12900.

>>101528320
No, no more kikery, Intel's response is probably bullshit anyway.

File: 1691728101944861.jpg (338 KB, 1200x900)

338 KB JPG

>>101528322
yes.

>>101527596
>previous cpu had the ebil vulnerability
>to update to latest directed version windows forced microcode update
>nuked performance by 25%
>many years later buy 13900ks for max performance
>sorry, you must patch it to nuke performance down to level of CPU that cost half as much
I’m finally done with Intel and moving on to AMD cpus. Will never buy an ATi GPU though

>>101528334
take your racism and shove it up your pimple, oily ass you inbred fucktard.

>>101528347
Shalom haverim

>>101528322
Considering programmers can't even make planes fly straight anymore I can completely see Intel who hires a ton of Indian men, the same Indian men who work for Boeing that cause planes to crash, to fuck up Intel CPU's as well.

>>101528347
Intel shills are shitty these days.
All the good ones were probably killed by Hamas.

>>101528371
Reverting to /pol/ talking points to mask away your inability to read a fucking article doesn't actually mask your retardation.

>>101528340
Windows microcode updates are easily disabled
I did it on a 9900k to train TSX

File: 1703366364600.png (42 KB, 408x666)

42 KB PNG

>>101528271
bad 13900K/14900Ks will request 1.5 V for 6 GHz
these are values from the VID table
the microcode was simply following the table, which contained "elevated voltage levels", and unless you think intel is brave enough to ship a undervolt to millions of customers they are going to be gating off these boost clocks

>>101528384
>read the article.
I did.
>Too much voltage
>Had to beat AMD somehow
>Released borked cpu
>Employed thousands of poos to defend intel
Repeat ad nauseam

>>101528347
Greetings to the jidf 4cha response team

>>101528405
None of that was in the article outside:
>Too much voltage
Which you are leaving out the fucking context.
>Intel's advisory says an erroneous CPU microcode is the root cause of the incessant instability issues. The microcode caused the CPU to request elevated voltage levels, resulting in the processor operating outside its safe boundaries.
>>101528399
I will not argue against bad silicon quality. That will be interesting to see. As I mentioned in a comment I left awhile ago here:
>>101527939
It would be nice to actually know what are the safe voltages. Again, I was giving an example with " 1.37 volt needed turning into a request of 1.47 volts because of firmware retardation" for retards here to actually understand what Intel was stating.

>>101528399
>VID is fused for 1.5v's
>firmware requests 1.6v's
>cpu dies
100% explains that supermicro buildzoid server 14900k.

They're clearly lying or people would've found this "fix" already.

>>101528478
they actually did months ago with cpu's that still were salvageable by dropping frequency to drop voltage.

>>101528477
that would have been noticed by people on day one
the CPU has a Vcore sensor that shows much voltage is actually being provided to the chip following Vdroop
the motherboard also has a sensor showing how much voltage is being provided by the VRMs

>>101527980
Replacing CPU's is cheaper than a class action lawsuit. They are trying to heavily optimize the VID to prevent or slow the degradation.

>>101528497
no one actually knew what the vids where. people figured their voltage tables were what they were seeing requested.

>>101528340
Crazy to think AMD didn't have problems either. This post stinks of AMDjeet advertising.

>>101528503
and seeing what was requested is what was being presented by the firmware.

then you get into the mess with ac/dc load lines being fucked around with by motherboard vendors. made it more tricky to see what the voltage actually was.

>>101528181
>No?
>Set your vcore to 1 volt and try for 6ghz
Look, you fucking idiot.
Yes, you do need minimum floor voltages to reach certain clock speeds, but you can also be higher than that voltage without further increasing clock speed.
This is why undervolting is a thing. The CPU makers set higher voltages than typically needed because of the variance in quality of the CPU dies, which helps with binning the chips.
So you can often (but not always) lower the voltages while maintaining the same clocks, and sometimes increasing the clock speed.

>>101528459
>understand what Intel was stating
Imagine believing in what Intel says.

File: IMG_1324.jpg (1.88 MB, 4032x3024)

1.88 MB JPG

>>101528503
>>101528512
open up HWiNFO and you can see this info right now
stop saying "oh we just don't know" because we know
both AMD and intel has Vcore sensors because it's hugely helpful for overclocking
>people figured their voltage tables were what they were seeing requested
the motherboard can show you the VID table

>>101528518
In this case if you lower the voltage you cannot maintain ADVERTISED speeds.
Which is bait and switch aka illegal in most countries.

>>101527596
don't care, i'm an amd supremacist

>>101528497
Voltages reported is displayed to you by the firmware. The firmware takes the voltages that were fused into the silicon, processes it, and displays it to you. Whatever you see, is what the firmware shows you. You don't really know what's actually in there because you can't read it yourself. You have to go through the firmware.

>>101528532
All of that is being displayed by the firmware. Not the actual fuse information.

>>101528543
>Whatever you see, is what the firmware shows you
the firmware shows you both VID (what the table contains) and Vcore (what the sensor is reading)
now why would intel mess with any of these? note that tech reviewers also determine power and voltage by directly measuring at the 12 V EPS cables
if the CPU was requesting 1.6 V someone would have noticed day one, and they definitely would have noticed after reports of degradation came in
>>101528555
where do you think the "actual fuse information" is contained?

>>101528704
Firmware does now show you the vcore. Vcore is showed by either the internal voltage regulator if the cpu has one (like haswell did) or by the motherboard. You can also obtain this by reading the voltage supplied to the CPU from the vrm rail on the motherboard.
Voltage table is fused into the silicon itself and the firmware within the cpu reads this information and presents it to you. Its reading is completely up to it. Its like taking another persons word for it.

>>101528704
and people were noticing 1.6 volts being requested months ago. the problem is, no one knew what the safe voltages were because the only thing intel ever stated in their documentation was "up to 1.7 volts" when alder lake came out.

File: Untitled.png (248 KB, 1024x768)

248 KB PNG

>>101528536
Yeah I fail to see how intel's solution would work
My 13900ks when new, only had ~30mv of undervolting headroom at 5.6ghz and not even 10mv of undervolting headroom at 6ghz

>>101527596
>BASED
... That isn't based. That's doing the bare minimum as a company that sold you a defective product.

>>101528734
They're in for a world of hurt. Let's just say that.
There'll be lawsuits out the ass.

>>101528740
Kek.
And they'll make you say thank you.
How magnanimous of them

>>101528720
>Voltage table is fused into the silicon itself and the firmware within the cpu reads this information and presents it to you
and if the voltage being requested was different from the VID in the firmware table, this would have been noticed by looking at Vcore, correct? are you with me so far or do you need this spoonfed to you again?
>>101528729
should be impossible unless you add a voltage offset
but stranger things have happened, for sure

>>101528734
It will magically work, trust intel

>>101527596
>"We tried to hardcore overclock our SKUs to win epenis benchmarks but it backfired!" "Thank you for playing QA and best-tester for our 7nm node!"

File: Untitled.png (27 KB, 862x314)

27 KB PNG

>>101528754
At least it seems intel's response seem to have appeased midwits like picrel
I guess the saga will continue in mid-August when people realize they've been had

>>101528243
It's not high trust. It's high volume. Not squabbling over blatant nigger returns and building those returns into your margins is the American way

>>101528773
the voltage wouldn't be presented differently to you.

if you read the article, the article states:
>Intel's advisory says an erroneous CPU microcode is the root cause of the incessant instability issues. The microcode caused the CPU to request elevated voltage levels, resulting in the processor operating outside its safe boundaries
that means the firmware was either miss reading (not translating correctly) the fused voltage tables or was reading it correctly but adding an addition to it due to a bug.

so what you saw requested would be was was drawn because all you can see is the request and what's currently being drawn. you cannot actually see the fused table itself. only the firmware can.

>>101528734
for example, that is the voltage table request. what the firmware is telling the cores to request. that is not the actual hardware fused voltage tables being shown.

>>101528788
>so what you saw requested would be was was drawn because all you can see is the request and what's currently being drawn. you cannot actually see the fused table itself. only the firmware can
lol so you're saying intel tested this feature for months, exposed it to motherboard makers and no one thought it was strange that 1.5 V was showing up in so many places
i'm sorry but this isn't possible, your 13900K won't magically do 6 GHz at 1.3 V

>>101528810
no, because intel's own released documentation to the public, since alder lake release, was "up to 1.7 volts for maximum vcore"
the problem, what i can now understand people are having a difficult time understanding, is that each processor released, has its own silicone grade.
one processor might be ok say with 1.34 volts, while another one might not be so. that's why each processor is shipped from factory with their own unique voltage table. no one processor shares the same voltage table.

>>101527596
>THANK YOU BASED INTEL
Yeah, thanks intel for solving a problem you yourself created, and also making us wait for it.

>>101528837
that means, yes, some 13900k's are fine at 1.5 volts to single core boosting for 56 seconds. others not so safe. if you have faulty firmware requesting to high of a voltage for that particular cpu, well you will have problems.

so if the hardware factory fused voltage table was post to be 1.47v's for that 13900k to single core boost to 5.8ghz for 56 seconds, but the firmware was buggy and instead making that single core request 1.51v's, that could fry that 13900k.

especially if c-states are disabled because then all the cores are getting hit by 1.51 volts instead of that one single core.

File: SP_rate.jpg (176 KB, 730x412)

176 KB JPG

>>101528837
>no, because intel's own released documentation to the public, since alder lake release, was "up to 1.7 volts for maximum vcore"
before it literally explodes
people always interpret that sort of technical documentation the wrong way
>is that each processor released, has its own silicone grade
the very best (0.1%) 13900K is programmed to request 1.4 V at 6 GHz
the very worst (0.1%) 13900K is programmed to request 1.5 V at 6 GHz
your average chips falls somewhere in-between, at 1.44-1.46 V

either way it's hell of a lot higher than a 12900K, and remember how 12th gen has no issues?

>>101528868
see >>101528865

So the server boards were juicing the CPUs with too much voltage?

>>101528870
oh lol so it's you, i already told you intel doesn't have "turbo durations" anymore
these voltages apply indefinitely
https://www.tomshardware.com/features/intel-shares-alder-lake-pricing-specs-and-gaming-performance/3
read, nigger

>>101528872
yes. buildzoid showed that perfectly with his supermicro minecraft server he leased.

it was shoving 1.5v's up to 1.6v's to run non stop, 24/7 5.7-6ghz for months

>>101528788
That would actually be retarded if it happened
iirc the fused voltages are communicated in a 4 or 8-bit hex address such that 0x0000 would mean 0.800v while 0xffff would mean the max vid specified by intel like 1.720v
But this system is decades old so screwing this up would be sheer incompetence

>>101528875
they do still have it. if you enable all the power saving features, including tau in the bias, which is now enabled by default with the intel baseline profile, one core will boost to 6ghz on a 14900k, and after 56 seconds, drop down to 5.7ghz, while another core will pick it up and boost to 6ghz.

it will repeat. you can watch this behavior with cinebench r20 running a single thread benchmark.

>>101528872
Yes, PL1 only limits power to the socket (125W). The CPU can still pull high voltage for ~6Ghz for a single/dual-thread boost. The game servers were hosting typically Minecraft which is single-threaded.

>>101528883
have you seen the indian men intel employees these days?

i actually live in chandler az. i see them every fucking day.

>>101528883
that anon has been coping for several days now
this shitshow has been so fucking funny
>>101528884
five seconds on google
https://www.tomshardware.com/pc-components/cpus/intel-issues-official-statement-on-core-k-series-crashes-stick-to-intels-official-power-profiles
the intel baseline profile is intended for bad motherboards, not bad CPUs, it won't do anything to stabilize your chip

>>101528903
that response has nothing to do with my post

>>101528868
Raptor Lake is just a tweaked Alder Lake with increased L2 cache on the P-cores build on the 7nm node.
Intel boost their higher-end SKUs into the danger zone in order to keep-up with Zen4 and Zen4 V-cache SKUs at gayming of all things. Gayming hasn't been much of an issue for CPU for years unless you are obsessive min-max type that frets over 300FPS+ and >0.03ms frametimes.

>>101528903
https://www.tomshardware.com/pc-components/cpus/intel-finally-announces-a-solution-for-cpu-crashing-errors-claims-elevated-voltages-are-the-root-cause-fix-coming-by-mid-august
>The company had previously advised its customers to stick with the basic power guidelines for its processors, rather than running them at fully unlocked settings, as it worked through the issues. Those instructions, which you can see here, remain in effect for now, and Intel hasn't issued any new workarounds for impacted customers. It is unclear if Intel will lift the existing restrictions after it issues the patch.
have you not read the OP's actual link?

>>101527596
I've been undervolting every CPU/GPU I've gotten in the past few years because of how ridiculous manufacturer's have gotten with their voltage curves, which probably saved my 13700k.

>>101528903
motherboard vendors default to the baseline power profiles now with the new bios. example, my asus defaulted to the extreme intel power profile for my 14900k.

even more funny, an article posted a day after yours: https://www.tomshardware.com/pc-components/motherboards/msis-motherboards-will-now-default-to-intels-official-power-specs-when-using-a-core-i9-to-prevent-crashing-official-power-profile-could-help-address-issues-as-intel-continues-to-investigate

shows msi doing the same as asus. gigabyte has done the same. they all have done the same.

File: SpaceTurtle_1-1719578483324.jpg (102 KB, 960x414)

102 KB JPG

>>101528907
>if you don't use stock settings, you get different behavior
obviously, but intel does not recommend that tau should be enabled by default
they don't want you to use it because it decreases performance and it won't fix the degradation
>>101528911
"intel baseline" isn't the basic power profile, it's a reduced power profile for motherboards that can't handle i9s
"intel default" is different (and they still recommend PL1=PL2, if the mobo can handle it)

>>101528911
nta he's not wrong
those settings wont mean jack for "bad" cpus that are already massively degraded and would only help marginally unstable cpus

>>101528932
> intel does not recommend that tau
they do, with the intel baseline profile they recommend it.
its why others like asus made another option, you can see in your bios, for dual tau to allow two threads to boost to 6ghz for 56 seconds instead of 1 thread.
>"intel baseline" isn't the basic power profile
it is the basic power profile. its intel defaults. they have two, performance, and extreme. its what all oems actually follow and now motherboard vendors instead of running everything at unlimited power settings and disabling things like c-states.

>>101528929
what a/c loadline does it use now? 1.10mOhm or 0.50mOhm? The former would be bad news because of way higher voltages leading to risk of degradation
This panic move by intel has been a total shitshow

>>101528941
my post had nothing to do about those settings helping broken cpu's. again, had nothing to do with my post

>>101528959
intel's ac/dc is still a "maximum not to exceed" which is 1.1/1.1 and both ac and dc have to be the same.

i don't know how other boards are doing it, but asus is finally linking ac and dc load lines together by default now. you do have an option to disable linking. at level 4 load line, my ac/dc is 1.0 / 1.0

>>101528926
This Raptor Lake debacle might be the end of that practice and it will start being locked behind "Do at your own risk" barrier. Just like how it was before Kaby Lake/Zen+ era.

>>101528977
1.1mOhm is only appropriate for worst-case A610 boards that have really bad socket impedence
The appropriate ACLL to set should be measured using the Intel VRTT tool for each individual motherboard that has its own unique vrm layout

>>101528979
and i hope so. amd does this.

>>101528996
intel honestly should set the requirements that vendors have to follow. instead of letting them do it themselves because up until now, they honestly didn't do shit themselves.

and seeing that supermicro board buildzoid showcased, they just set it to 1.1/1.1 and called it a day

>>101528996
So if asus' layout has 0.5mOhm of resistance, setting it to 1.10mOhm default would cause an overvoltage of 0.60mOhm, which would be upto 0.120v (?) of overvolting under extremely heavy loads and 0.03v (?) of overvolting under light loads

>>101528957
i was going to say "intel extreme has no tau" but whatever, i don't care anymore
this shit is confusing and broken for no reason

i think it's evidently clear that intel needs to throw out their broken decades-old system and implement a modern opportunistic boost system with wide safety margins like AMD does
concepts like "turbo durations" and "per-core ratios" shouldn't exist on a modern CPU
until they fix that i see no reason to trust any new desktop chip from intel, they can add half-assed workarounds to limit degradation at the cost of performance but i don't trust them to do that since they already went ahead with 13th gen and 14th gen

>>101529015
intel extreme does have tau.

>>101529025
ok

>>101529031
all the extreme profile does is allow for higher all core frequencies compared to performance profile. both share the same single core boosting behavior as both have enough headroom for single core work loads with their allocated wattage / amperage tdp values.

>>101529040
ok
your intel CPU has AIDS

>>101529015
people have been asking intel to set actual requirements for over a decade. this shit has been going on since 2009 when "mutlicore enhancement" hit the scene. and its only gotten worse since.

>>101529054
you don't understand, a modern CPU should be able to protect itself
"tau" or "power delivery profiles" shouldn't be necessary
you can set unlimited PBO limits on AMD and as long as you don't touch the scalar the chip will not degrade, the limits are intended to protect the motherboard only

>>101528884
That explains why Buildzoid was saying "24/7 5.7ghz and 6ghz single for months at 1.5 to 1.6vs" I was confused but it wouldn't have just been 6ghz. That makes sense because it was jumping back and forth after X amount of time.

>>101529079
Yeah, that's because AMD set a limit to PBO for end users didn't kill themselves and AMD had to issue a warranty claim.

Intel didn't because they have been retarded about this since 2009.

>>101529090
>that's because AMD set a limit to PBO for end users didn't kill themselves and AMD had to issue a warranty claim
you can set all PBO limits to 9999 and the chip will not degrade
safety is controlled by an internal "FIT limiter" which takes dozens of variables into account
you can decrease the safety margins with the "PBO scalar" to get higher overclocks, but this isn't what i'm talking about

they don't allow it PBO by default because it would make their power efficiency look worse, remember how the 5950X looked really fucking good in efficiency compared to every other chip out there? that's because it was hitting the PBO limits

>>101529090
Also, PBO is locked and requires end to activate and gives a explicit warnings that activating it may result in data loss, irrevocable damage to hardware and AMD isn't responsible for it. The end user is.

>>101528405
>amd retard cares more about defending team red than truthful information
why do they even hire people like you to defend them online when they could just train chatbots to regurgitate all the same talking points ad infinitum?

File: 1719980094522940.gif (119 KB, 419x281)

119 KB GIF

>never fell for the estrogen cores meme
>jumped ship to 5800X after my 10700k
LTSChads just cant stop winning

>>101527654
Amazon will just send it to a new customer and the new customer's gonna be like "WTF amazon why do my CPU not work? and Amazon's going to be like tough shit."
Fuck Amazon.

>>101529164
This is why you always claim defective for reason of returning. They send it straight to either ship back to vendor or liquidate it.
Customers that get previously returned items get it because the originally person returned it under another return reasoning like "I didn't want it anymore, changed my mind, ordered by mistake, etc."

>>101529090
thats because intel for a long time had the blessing of headroom with their designs. a 2500k could, without effort, be overclock a minimum of a full 1000mhz. they didn't need to worry about motherboard manufacturers fucking around with shit like multi core enhancement because of that headroom. raptorlake has eliminated that.

>>101528996
Now what is the purpose of AC Loadline? SIMPLE
It's SUPPOSED to boost the operating voltage by PREDICTED CURRENT.
So if your CPU has a 1.30v VID at 5.4 ghz and you are pulling 100 amps of current, with a 1.1 mohm AC Loadline, it should boost the operating voltage by 100 * 1.1=110mv, so the operating voltage BEFORE VDROOP will be 1.410v.

>>101529487
Imagine you have one of those lemons that already have lets say 1.500 V. And you push only at around 50 amps into them

1500mv + (1.1 * 307) - (1.1 * 50) + 0 =
1500mv + (337) - (55) =
1837mv - 55
1782mv = 1.782v

>>101527596
>>101527634
>just keep your PC turned off until August goyim

>>101529508
With little to no headroom since 13th gen, Intel really fucked themselves hard. They had to be extremely precised now instead of sitting back and enjoying the headroom to fill in their dumbfuckery.

>>101527596
how do overclockers know what the safe voltage ranges are? the bios just has 1000 buttons to adjust them but nothing anywhere says clearly what the ranges are

>>101529158
>talk shit about estrogen
>post estrogen pic
.

File: 1717913062459257.jpg (75 KB, 1080x1080)

75 KB JPG

>>101529570

>>101529551
trial and error. but with raptor lake, no one really fucking knows because they are pushed to their limit at stock.

>>101529487
>>101529508
>>101529534
These numbers and intel's claims don't add up
I think they're lying to cover their asses

>>101529584
sounds pretty dangerous. i don't even have the slightest idea what the massive amount of fine tuning values are or do. back in the days this all was much easier

>>101529588
this wouldn't be an issue if intel had actual headroom. with 13th gen pushing frequency up hard and 14th gen continuing it, problem is, for yields, they had to let A LOT OF FUCKING LEMONS through. the "best" silicone has little headroom but enough for most to get by, and the lemons have none. for those lemons to remain stable, EVERY FUCKING THING has to be set just right or they will start dropping like flies, which they are.
with mobo vendors at first running shit with zero limits, unlimited tdp, and fuckery ac/dc load lines, they were just killing lemons left and right. now with the new power profiles it reduces it, but the issue with LIGHT loads are still there. profiles just help reduce failures with heavy loads (which ironically is better because they DROP voltage to stay within the new enforce power limits).

File: 1687988928546375.png (65 KB, 225x350)

65 KB PNG

I've been asking (((pcbg))) for the last 2 weeks
i9-10900T ES QTB0 1.5GHz 10C 35W For H410 B460 Z490 LGA1200

can someone please tell me if I am retarded or the world is for pricing this shit at 130 dollars?

>>101529642
>heavy loads (which ironically is better because they DROP voltage to stay within the new enforce power limits).
what i mean by this, is that ironically enough, you are safer running a heavy load than a light load. if you play games that leverage a lot of threads you ironically will be running at lower voltage than playing a game that only stresses a single or two threads. all because with reduced tdp, frequency will have to drop, which lowers voltage.

>>101529642
how can they even afford that... i head amd has basically no trash parts

File: 1716413225715691.jpg (41 KB, 476x477)

41 KB JPG

>>101529657
>130 dollars for an engineering sample that maxes out at 35 watts
pic related is your answer.

>amd
lol
https://old.reddit.com/r/AMDHelp/search/?q=crash&sort=new&restrict_sr=on

>>101529678
well they couldn't afford it which is why they let a lot of lemons through. if they could afford it they wouldn't have needed all the "trash parts." this new update intel will be releasing will probably be cracking down more on mobo vendors fucking with voltages and i wouldn't be surprised if a lot of 13th / 14th ends up having lower clocks with lighter loads. all core sustained probably won't see a change, but single core sustained will probably be nerfed a bit.
amd really won with chiplet designs and not trying to push their shit to the limit.

>>101527596
OI vei goim, turns out the ROOT cause was ANTISEMITISM all along

File: 1719030415909628.jpg (2.38 MB, 4032x3024)

2.38 MB JPG

>>101529635
like what are these things why is there more than one voltage for the cpu? celerons and pentiums didn't have anything like this

File: why are you doing this.jpg (156 KB, 1920x1080)

156 KB JPG

>>101529680
i dont want a higher wattage cpu in a case as small as I'm putting it in but by your look I can assume it's still a ripoff. what is actually a good price?

>>101529742
for something that old and that nerfed, it shouldn't be more than $25.

>>101529508
Good lord Intel what the fuck. It explains why their documentation lists 1.7v's as the max. Its not the max to set, but the max because of transient spikes. Problem is so many trash parts got released and oem's not setting AC / DC load lines correctly even though Intel sent them the fucking tool to do it is causing so many parts to spike PAST 1.7v's with transients.

>>101527596
>>However, the microcode update will not repair impacted processors.
lmao

>>101529800
>software update can't fix a bullet hole in a cpu
>LMAO INTEL SUCKS

>>101529802
Both greentexts are true, i agree

>>101529802
>no, recall and replacement isn't a solution. here have a software patch if it make you feel better bro
nigger

>>101529833
they are doing replacements

>>101529678
They have trash parts like the ryzen 7 5700 (MUCH slower than the 5700x since its an apu with igpu disabled), their XT lineup of cpus and weird mobile naming scheme to obscure which generation they are (7730U having worse singlethreaded perf than 7640U etc.)

>>101527738
lol that will only "work" until it suddenly doesn't one day and all your previous transaction are now up for audit
good luck

>>101529849
>all your previous transaction are now up for audit
where the fuck do you live where they do shit like that?
>hurr walmart will randomly issue an audit for all your transactions!
in what fucking universe? you live in europoor or something?

File: 1718517859158902.jpg (328 KB, 2048x1371)

328 KB JPG

>>101529775
alright, sounds like i'll have a hell of a time finding one but I wont compromise.

>>101528181
>tard
Boosting voltages is an easy way to help stability at higher clock speeds but it isn't a requirement.

>>101529657
Don’t do it. I wasted years with an engineering sample Skylake i7 6400t. Degraded fast and could not hit 4ghz after a while. And that i9 probably cannot BCLK OC like my old 6400t could.

>want gaming CPU?
Get 7800x3D or 5800x3D if budget is lower
>Need cores?
Get 7900x/7950x
>Need only IPC?
Get 7700x or 7600x
It makes sense for data centres to buy Intel CPUs because they get different prices and services but why are consumers buying these 250W hot garbage?

File: yes old good and new bad.png (977 KB, 1484x1643)

977 KB PNG

>>101529957
im putting this in a pizza box case if anything I'm going to power limit the voltage, I just want an i9 and to max out the slot if possible, but I dont see any reason to pay 400 dollars for speeds I wont use. my i3 is already snappy enough, but I'd like to have the capacity to run background tasks, but spending 400 dollars on that? not gonna happen. even 130 is too fucking high.

why do people think these things are worth so much? I just want to max out a MB almost 5 years old and still using pcie 3

>>101530029
there has got to be SOME competition around so we can buy our cheap x3d cpus my friend

>>101530086
x3d literally burns holes into sockets and has a higher failure rate than their non x3d counterparks by amds own admission

>>101531154
Update your script
It was fixed within a month and all affected cpus are covered by warranty

>>101531172
nope, it can never be fixed because x3d is unstable as shit and runs hotter. its why amd has to bin hard for low tdp chiplets but you're a raging retarded byproduct of incest.
i would have an easier time arguing with a wall than you stupid amdrones

>>101531187
>amd drones
if you love intel so much sell me a i9 10900t for 40 bucks.

File: 1720269852242839.png (757 KB, 674x1200)

757 KB PNG

>>101531187

>>101527654
This lol

>>101531209
>nooo why won't you love my pile of shit
when will you amdrones realize not only do people have extreme hatred for your shitty brand but people would rather eat actual shit than even get near let alone touch your vile ryzen shit? amd is so shit its more filthy than actually shit that comes out of your pathetic ass.

File: 1528911053995.png (196 KB, 807x745)

196 KB PNG

>>101531255

File: Screenshot_20240723-010013.png (407 KB, 1080x2400)

407 KB PNG

>>101531262
>nooo you are coping if you don't enjoy 4 minute boot times or 11 second boot times + increase in BSOD's
againt take your shit ryzen and shove it up your obese ass you fucking freakazoid

File: 1575552938959.png (207 KB, 800x612)

207 KB PNG

>>101531302
>dual rank 2x32gb 6000cl30
>21.3s bootup
werks on my machine

>>101531302
>MSI
What's next you're going to say biostar has issues?

File: Screenshot_20240723-010457.png (354 KB, 1080x2400)

354 KB PNG

>>101531302
>lol you want slow boot times instead of dog shit slow boot times?
>enjoy your bsod and corruption
>intel just can't compete
i fucking hate you amdrones so much its fucking unreal. all you vile filth do is FUCKING LIE ALL THE FUCKING TIME

File: 1754716014617950.png (926 KB, 801x1500)

926 KB PNG

>>101531343

>>101531321
>amdrones proud to take 20 fucking seconds + data corruption is "just werks on my machine
my god I can't fucking stand you peices of shiit. LIE LIE LIE MANIPULATION IS ALL YOU AMDRONES DO

after you fickwits burned me by convincing me to get that dumpster fire 8350 instead of a 2500k i vowed to make your pathetic lives as miserable as i could for the suffering you caused me by that piece of shit faildozer

>>101531368
What do you mean, I just spent 6 hours overclocking my FX-8350 yesterday and had a lot of fun

They claim that this microcode only effects desktop chips, but is it possible the laptop chips are pulling excess voltage too? I've had strange issues lately.

>>101529802
This is the dumbest cope I've ever seen. Intel caused the "bullet hole", you dumb faggot.

File: 1496622518440.png (973 KB, 801x1500)

973 KB PNG

>>101531366
these images are so fucking funny
can someone post the one with the "bing bus" lmaoo

>>101531366
>>101531432
>can another amdrone come in and stroke my small pp in a circle jerk plz

>>101531424
>amdrone actually thinks a software update can fix a bullet hole in the head

>>101531478
>doesn't understand the point
I get that you have to be dumb to buy Intel, but I didn't realize THAT dumb.

i bought athlon in 2005 and c2d in 2007 and phenom in 2009 and i5 3rd gen in 2014 and zen3 in 2022 as these were the best deals at the time. i dont care for your stupid brand wars please kill yourselves

unfortunately due to this latest intel debacle im going to have to break tradition and go amd twice in a row soon

>>101531368
>8350 instead of a 2500k
Based dumbass

File: 1615037593636.png (876 KB, 801x1500)

876 KB PNG

>>101531432
Kek here's the 2020 update

>>101527869
> i did this with my last msi motherboard. i broke its pci-express latch and got some thermal paste in the am4 socket. so i bought a new one and sent my old one back for the refund.
Fucking scum

File: 1721726973482.jpg (841 KB, 1536x2048)

841 KB JPG

Too fucking late, this bad boy ia ready for assembly

>>101531931
what do you need 4 ram slots/3 pcie slots for?

>>101527679

> Intel won't just tell you you're out of warranty so take them to court or gfy.

Intel will just wait for the class action suit to wrap up in 5 years, of which consumers will get a 10% off a new CPU coupon.

>>101531944
Filling them up wirh ram sticks and ssd drives

>>101531187
my 7800x3d maxes out at 79C during linux compile, with undersized cooler in an mitx case
keep coping, and remember to rma youre cpu every couple months

I wonder how the benchmarks will look like once it's "fixed". Can't wait for GN video with details from FA lab.

>>101527703
based
brb buying all intel processors in stock

>>101531872
Kek

i decided to never buy intel again since they patched away my i5 2500k clock speeds 10 years ago. i made that decision out of spite but i guess it has turned out to be a regular thing that they just do now lol

>>101532040
>consumers will get a 10% off a new CPU coupon
$5 at most

>>101532073
>I wonder how the benchmarks will look like once it's "fixed".
Why do you think the microcode release is targeted for after zen 5 launches lmao. It's going to be another spectre tier impact.

>>101527726
sellers?
its shipped and sold by amazon.
amazon is the seller.
do you think they care? they rma it from intel and intel eats the cost, not them.
if intel tries coming back claiming the serial is out of warranty, Amazon will hunt down and shoot another intel worker as a threat.

>>101531597
based good product enjoyer anon

File: 1721750496396.jpg (71 KB, 602x727)

71 KB JPG

>>101529185
not my problem. lol.

>>101533479
the spectre, meltdown, zombieload patches during 2018-2022 were brutal
7700k's, 8700k's and 9900k's lost between 10-80% of their performance depending on the workload

>>101535419
>all that money and energy wasted on hypothetical, unlikely attack scenarios

>>101528536
This sounds like in edge conditions the CPU would request higher voltages than necessary in certain operating conditions due to a bug.

I'm speculating, but imagine you are interpolating along some sort of polynomial in integer math (we are talking microcode here) and you make a mistake in your calculations maybe due to round off and you get a 0xFFFFFFFF voltage request output to your vrm - but only during edge conditions This could be hard as fuck to track down

File: 1498458774248.jpg (324 KB, 882x758)

324 KB JPG

I see the intel shills are coping pretty hard.

>>101531931
>erp ready

>>101527697

Amazon will not look in the boxes, as long as there's a CPU the nigger that checks if the box has a processor inside it flags it as correcly returned and the problem becomes the next buyer's problem.

>>101531255
>>101531302
>>101531343
>>101531368
>>101531475
>>101531478
Why is UserBenchmark posting on 4chan?

File: GTIlMXlawAAVTB-.jpg (70 KB, 941x1061)

70 KB JPG

>>101527596
>>>MID AUGUST<<<
>"THANK YOU BASED INTEL"

File: rplel.png (72 KB, 1232x411)

72 KB PNG

>>101527596
But they havent found the root cause
What they found was A root cause

>>101527596
Faulty internal voltage sensors and usage of impure chemicals (synthesized in India) used during manufacturing. Something like that will be the root cause

File: Intel.png (522 KB, 1247x1266)

522 KB PNG

>>101546629

File: Tranny.gif (2.29 MB, 251x255)

2.29 MB GIF

>>101546866
inspect element but wouldn't be surprised if it was true either ;p

>>101546511
The root case is hardcore overclocking at the factory, because Intel marketing wants to Intel to #1 no matter the opportunity cost.

>>101546866
Sirs…