According to specs on the official AMD web side, current generation desktop CPUs supports up to DDR5-5600. They are not expected to make 6000.

​@@soonts Is that why they keep telling reviewers that 6000 is the sweet spot and that they should use 6000 ram for every single test they are expected to make?

@@firstnamelastname366 I agree it’s unfortunate AMD tells different stories to reviewers and consumers. In ideal world, I would prefer reviewers to ignore everything vendors tell to them, and only listen to what vendors tell to consumers. In the meantime, consumers have an easy workaround: read CPU specs, and buy standard memory of the supported speed which doesn’t rely on Expo / XMP to deliver the performance. As a nice bonus, old-school green memory like Kingston ValueRAM is slightly cheaper than gamer-targeted stuff with heat sinks and LEDs.

LOL.. well in all fairness that's not all true, because if you got it from Amazon/Newegg in the US or Canada, you can send it back for replacement as defective. Elsewhere without free/open returns is more of the issue.

@@firstnamelastname366 Jay's 2 Cents got feedup of the AMD memory speeds issues and nerfed Intel by running their CPUs bellow memory spec in some tests a while back. 🤣 Back in the day, reviewers had Stock and OC results. Why can't they go back to that. OCing the memory controller might be less of a lottery draw but it still is.

That is a catch-22 Here in the U.S. almost all electronics are sold with a clause that from FCC : 1) This device may not cause harmful interference, and (2) this device must accept any interference received, including interference that may cause undesired operation. FCC Caution: Any changes or modifications not expressly approved by the party responsible for compliance could void the user’s authority to operate this equipment. (So things like putting a set of car keys in a faraday cage/box to protect from car thieves using a signal booster/relay...is technically against the FCC complaince in the producty warranty However I do not think it gets enforced with such products markets for sale (ie...faraday box for car keys) Same for RF ID blocking credit card and passport covers/books that aren't FCC certified. FCC regulations require the clause in the warranty of the product...yet it doesn't seem to be enforceable for warranty claim denial.

@mariano3113 where would FCC require it. That's all just the same bull that Linus said about warranty

Just lock the chips. They did it with xeons. Zero ram overclocking. The IMC simply ignores requests to run faster

It may be interesting for you to know that X3D CPUs are much less significantly affected by slower memory speeds, at least in games, so they can actually still run almost as fast in games and some other applications even without using XMP profiles or any form of memory overclocking. This could actually make them even better than other CPUs for some types of non-gaming workloads as well, because the higher stability of running the memory at the official specs is more important for some types of non-gaming workloads.

They never advertised the cpus in the literature that they can do 6000, it says right on the product page 5200. They say the "sweet spot" is 6000 cl30, not that 6000cl30 is the official support. While I agree its misleading to have reviewers benchmark with these "overclocks", legally, they can deny the claim if the IMC cant hit 6000 1:1

None of that is accurate, except the part where they can deny the claim, but that has nothing to do with whether it's actually legal.

That's true. However, the XMP/ Expo RAM speeds do not meet the same standards in terms of error rates. If you want your system to run as stable as possible, and to keep occasional errors at a minimum, you should run at the official supported memory speeds. The memory and the CPUs have been thoroughly tested at a range of different memory frequencies, and the official supported speeds are based on meeting a certain threshold of stability at that speed. For gaming, the errors aren't common enough with Expo enabled to be a significant enough problem to worry about, but in some types of professional workloads, it can be better to run the memory at a lower speed. With XMP on Intel 13th/14th gen CPUs, I'm not so sure if it's safe to run with XMP enabled even for games, but it probably is, at least if you have the latest bios updates installed. With XMP for Intel 13th and 14th gen CPUs I don't know if there's any strong evidence that the degredation problems can really be blamed on running memory with XMP enabled. However, if the degradation problems have affected a 13th or 14th gen CPU, running the memory at a lower speed like 4400 or 3600 does often allow the CPU to still run stable again even after the chips had suffered permanent degradation... which makes me think that not using XMP may have had a significant effect on the chances of the CPU suffering permanent degredation in the first place, even if it seemingly isn't the direct cause of the degredation, and this might still be true even after Intel's updates which are supposed to prevent the problem from manifesting in chips which haven't degraded yet, but, in my opinion, that doesn't seem particularly likely.

Where do you guys even come up wih this nonsense? I've been building and repairing PCs professionally for over thirty years. SIPs, SIMMs, DIMMs, RDRAM, ECC, you name it. No errors are tolerated, none, at XMP / EXPO or the offending component goes back. No "occasional errors", no JEDEC speed, and the Raptor Lake defect has exactly zero to do with XMP / EXPO. These companies CYA on product pages while literally advertising the higher speeds in nearly all marketing material, which most of you never see as consumers, and you just eat it up, too funny.

"And the memory makers can produce modules that run that speed... " There is hardly ANY DDR5 >5000 available. Those sticks advertised as 8400? Yeah that is 4800 and then overclocked by increasing the voltage to insane levels. And i mean insane levels - often it is over 250mV.

Mobos (and the Taiwanese ODM in general) have and always will be the #1 offender. They live in the wild west (especially Intel because Intel does a pennyante job at creating and enforcing standardized specs) where they can make BS claims, offer/set settings that fry hardware (they all HAVE and CONTINUE to do this), refuse to spend the extra couple bucks and minutes in manufacturing techniques to make sure their boards can run hardware correctly, and then refuse RMA if your board has one spec of dust or fingerprint on it. This is also why I find it amusing how people were bellyaching over Zen 5 not having a new IMC. If AMD did that, besides Zen 5 needing to get delayed into next year, then there would also be endless issues and headaches for AMD, because most AM5 boards are not fit to handle what a new AM5 controller could do and everyone would take it out on AMD because the mobo ODMs are always exempted. They similarly botch Radeon (and GeForce) cards all the time, putting low-grade hardware, inappropriately small coolers and inexcusably cheap paste and pads, and then people have issues that are not technically the fault of the GPU hardware. People need to stand up to the Taiwanese giants because AMD can't do that alone. And Intel doesn't care what anyone does until it causes PR nightmares for them (when they'll throw the ODMs under the bus regardless), so that's probably how they built up that attitude to begin with...

Doesn't matter, if the MOBO advertises support for it, if the CPU maker not only advertises supporting it - also all their marketing benchmarks are run with it -, and the RAM is sold advertising that speed... It must do that speed. Otherwise it's false advertising, a literal scam, since you pay more to access that performance

I'm running XMP on my 7800X3D. Works just fine. 96GB 6000MT CL30.

@zach1860 All memory has JEDEC and EXPO / XMP profiles if it's beyond JEDEC. The EXPO / XMP profile has separate timings and voltage requirements. Most boards adhere to what the profile specifies, though some do not. You have to check in the BIOS or an application to see if your motherboard matches what the RAM advertises. If a board or CPU is particularly weak / binned, or the BIOS is absurd about settings, yes damage is technically possible, but very unlikely.

Depends on your motherboard

Defective CPU, request warranty. That's exactly what HUB-guys are saying: You will be covered by manufacturer. Otherwise - contact Hardware Unboxed / Gamers Nexus with message history between you and Intel/AMD; and wait for some juicy content.

Did you check the motherboard ram compatibility list and was that ram in the list as compatibal? If it wasn't then it was you at fault not anyone else.

Bro, have you even checked the compatibility of the motherboard? I never had this issue with my z690 Mobo with XMP enabled at advertised speeds... Just because your CPUs support those speeds doesn't mean that your motherboard will

6 layer PCB posting

​​@@randysalsman6992 Why would a non supported ram have an xmp profile on that motherboard.

That sounds like your 13600k may have suffered from the permanent degredation problem which is now widely known. This problem could happen even if you ran the memory at the official specs, and running your CPU with XMP enabled should not void the warranty. Look into testing your CPU to see if has been affected by this degredation problem, and look into filing a warranty replacement claim for it.

Your feelings still don't change the reality of things.

He's voicing his concerns regarding the moral ambiguity of calling XMP / EXPO overclocking. Especially when AMD / INTEL show their performance based on XMP / EXPO results. Personally I wouldn't call these PROFILES overclocking, more fine tuning IAW your hardware. Be better ​@@randysalsman6992

You're not OCing the ram, you're OCing the cpu

You sound like someone who would buy DDR5 and cry because it doesn't work on a DDR4 Mainboard.🤣🤣

​@@alberts4541 DDR5 on a DDR5 motherboard seems like a stretch ... If it was said " type of person that would use a PCIE Gen 5/Gen 4 drive in a PCIE Gen 3 and complain about the performance" that might makes more sense in this circumstance. -Slotting RAM that can work at different speeds in the same compatible slot...similar to M.2 drive physically compatible to slot in, but wrong Gen or shared lanes with Sata/etcetera results in less than advertised speeds.

well the ram supports it, doesn't mean whatever cpu you pair it with will.

Who is saying that any warranty claims are being denied because XMP or Expo was used? Where is the evidence?

@@syncmonism nobody said it was being denied , can you even read? did you even watch the damn video ???? , this is about statements and claims they made themselves even if they dont follow their own words , how fcking dum are you? btw why are you trying to defend a company that dont give 2 f's about you and only want your money ?

@@SpecialEllio hello another dum f why are your bringing this up ? why are you bringing up cpu ? we are talking about warranty claims and you come with cpu compatibility??? your comment has ZERO to do with what my comment was about or the video was about ? do you live in your own little reality ?

do you 2 even have a brain cell or the capability's to understand english and what this video and my comment was about ?? BRAIN ROT DETECTED?????

so is the AGESA allowing shoving 2V into your CPU...

@@PSXman9 sure but agesa never gets advertised unlike expo or xmp

they even got QVL LIST in their motherboard basically telling people to buy those Overpriced RAM. just to hit those advertised speed.

It can though, because to run that EXPO/XMP enabled, high speed kit you also need to run UMC/IF or SA/Ring at higher clocks too, which is out of spec and is considered overclocking and overclocking legally voids your warranty.

They are not denying warranty requests for people who used Expo, or at least AMD isn't. Intel are the ones who have been having very high failure rates, but I have yet to hear about them ever denying a warranty claim because someone had been using an XMP profile.

CPUs don't overclock themselves and voltages are controlled by motherboards... Intel CPUs were too close to their max point out of the box. My 13700k is going hard after 2 years because I disabled all overclocks on BIOS applied by default by the Motherboards. Matter of facts all the fixes are all motherboards setting fix not CPU... So many ignorant people baited by these "highly informed" KZbinrs... They told you 2 years ago about these problem. Instead of pushing for a change they stayed quite and now they are act high and mighty like they are interested in the consumers when in reality they are happy to have content to talk about... Der8auer said it that stock settings applies by Mobo weren't from Intel, he said it 3 years ago when 13th gen dropped...

Car makers set a 'default' tune. CPU makers set a 'default' tune. RAM makers do. Every machine maker does to some degree. Most processor i7 and i9 units from the past 2 (not three, you lost the plot here) are fine. The amount affected is like 1% or less. The first time Intel has had instability of any kind in like 20 years and it's extremely minor, already patched and fixed, and affected a tiny tiny amount of units negatively. You don't grasp the whole situation and are being a chicken little.

@@jeanytpremium ~ Yeah, fair criticism. I am not defending the tech tubers, your call on them is exactly right. I should perhaps add my personal hobby-horse. Cut n paste from an earlier comment ~ ----- Personally, I think the point is more about having marketing and bookkeeping people, boardroom pirates, running the company, not engineers. If you are a builder, you have a builder running the company. If you're a shipyard, you have a naval architect. If you're a chip design and fab company, you have a fabrication and computation engineer running the company, not a marketing guy, not an accountant, not a boardroom stock-options pirate. Now they are trying to fix this with Pat Gelsinger, who did once work at Intel as an engineer. But it's largely too little and too late. Lisa Su was brought into AMD when they were at a low point, caused by almost the same problems, and she has saved them. Now it's up to Pat to save Intel. So far ~ he is making the right noises and seems to be moving in the right direction, but [sigh] I don't know if they're going to pull out of this dive or not. They are still not in quite the same fix that AMD was when Lisa arrived, but they're in a lot of trouble. I did say, in 2016 or so, when news of Spectre & Meltdown arrived, that Intel was doomed. I said it might take 10 years to play out, Intel are a rich and deeply established company, but that company is going down. Looks like one of my better & more accurate predictions. That was before we had the 12 ~ 13 ~ 14 gen burning themselves out and getting silicon degradation because of over-clocking and over-volting. Now we all get to watch and listen to how they deal with that. -------- I blame boardroom pirates, the manager class who know stock options but don't know much about about CPU design or fab workings.

@@CyberneticArgumentCreator ~ Maybe. Cluck cluck. See my other reply on this thread. Intel are far from dead ~ but they need to turn a whole lot of stuff around, and they don't have infinite cash rolled up in a sock to see them through. In support of your point, looking at the other side of the coin, we heard about Spectre & Meltdown and a million side-channel attacks that Intel was wide open to. Ok ~ some years later ~ we do have performance robbing mitigations in the Windows kernel and the Linux kernel, and we do still have new permutations of that stuff come up once a week, but has anybody ever had or seen a report of Spectre or Meltdown or any of those tricks, being used in the wild, by bad guys, to do bad things? Because I haven't seen one. And I have been watching. Cluck cluck. Mark Twain : "The rumours of my death have been somewhat exaggerated."

@@CyberneticArgumentCreator The issue is that Intel didn't set a default tune and then blamed the motherboard manufacturers. Gamers Nexus talks a lot about this. Intel has not in years set a "this is the default spec. Anything over this is on you." They've been giving a sort of vague suggestion of a range of settings. Probably just so they could shift blame to the MB manufacturers when something eventually goes wrong. Even with their "fixed Intel settings" they gave different motherboard manufacturers different numbers.