Can you explain to me, please, I beg you, in what workloads does an Intel pulling 300w lose to a much lower clocked CPU that's almost a third the power usage? I'm really dying to know. As far as I understand, you are talking about the 7800x 3d, which gets absolutely blasted in any workload that the Intel chips draw 300w. It's nowhere - absolutely nowhere near. So what are you even talking about? Why do we have to keep reading these ***tposts that have no basis in reality? Just stop it, please.
If we are talking about CBR23 (which is a common metric), a completely unrestricted 14900k gets 42k (with no thermal throttling and max priority). A 13900k gets a little bit over 41k. With a 253w power limit the scores are 37 and 36k respectively. The 14900ks might score a bit higher, but I don't think it can hit 38k. Maybe 37.5k, but haven't tested one myself. The other 2 I have.
That's with AC/DC calibrated properly for accurate power measurements. I've seen people on reddit claiming that they can hit 39k at 253w. No - they can't, it's just that their AC settings are missreporting powerdraw.
craigss
Reputable
- Aug 28, 2020
- 54
- 36
- 4,560
Ahhh remember our conversation?
It seems Intels quest to be at the top of all the charts irrespective (remember the 10900k nonsense) has come back and taken a bite, bit like the time they demoed a system that had a hidden cooler on a live stage and declared it was just a CPU.
Intel has been proven to cheat lie and blackmail its way to the top of the charts already why are we all surprised that once again they have been caught red handed, sure all those settings were fine and dandy .....until trouble came calling then its suddenly a Motherboard issue.
In reality its a greed issue Intel just wants to get to the top of the charts no matter what and the child Gelsinger is the main driver in all this as he was last time.
It has nothing to do with Intel. Reviewers (eg. HWunboxed) have admitted themselves that they go into the bios to enable XMP and they specifically choose the "power unlimited" option. You do know that in most - if not all - Intel motherboards, you can't even enable XMP before choosing your power limits, right? The first time you boot into the bios there is a huge image taking up the whole of your screen asking you what power limits you want, usually there are 3 options, something like 150w / 250w/ unlimited (it depends on the motherboard, but that's pretty much it). It's not an Intel problem if reviewers choose the unlimited option (and then they measure efficiency with that option, lol).
craigss
Reputable
- Aug 28, 2020
- 54
- 36
- 4,560
Another Fanboi yada yada yada
Intel has been caught up in this becasue Nvidia rightly pointed the finger at them and now Intel is trying to offload the blame to the boards who incidentally must pass Intel vetting before release, whilst there were no issues Intel was complicit in allowing all manner of Bios changes to make themselves look better, how many reviews have we all seen when the CPU say a 14900ks was drawing almost and in some cases over 400w, De Bauer has a review online, if when those reviews came out it was an issue for so much power Intel at that point should have stepped in to stop it, in fact at the validation point they should have capped it but no becasue it got them a few extra Hz they let it go until suddenly its an issue because lets face it the ks or even the top end k CPUs are sold as premium top of the tree balls to the wall products that are capable of overclocking indeed they sell them as such through board partners who unlock stuff that frankly results in stability issues as we are seeing.
For the record the highest consumption I have seen on a none LN2 platform was 508w on the Tech powerup site.
At that point Intel had a duty to step in but they didnt as it allowed them to wave the Epeen, now reddit is full of users complaining of stability and chip degradation issues nice job Intel.
TerryLaze
Polypheme
- May 9, 2016
- 9,996
- 2,695
- 63,440
If every electronic circuit can be burnt by pumping SOC voltage out of spec then why don't intel CPUs blow up the same way that amd cpus are?!"Intel just crashing is still much more competitive & better for the end user" Is this English ??? That Gamernexus video just showed CPU can burn by pumping SOC voltage out of spec. It doesn't mean anything because any electronic circuit can be smoked that way. If you're not a bot, your reasoning is adorable i must admit
Intel CPUs get pumped not only with volts but with watts and ampere as well and still don't blow up, they do crash just like every CPU in the last 30 + years has had crashing at extremely high overclocks.
Do you remember this current conversation?!
AMDs quest to reach at least close to intel forced them to release a whole generation of CPUs that blew up under default out of the box settings.
craigss
Reputable
- Aug 28, 2020
- 54
- 36
- 4,560
Actually they didnt the user was using an overclock setting and if I recall AMD wanted to replace the CPU and have it back for inspection BUT a certain weasel on you tube who thinks he is uber important threw money at the owner so he could sensationalise it, do AMD cpus fail? Yes of course they do all products have a failure rate never said they didnt and frankly this whole attempt by you to once again divert attention from the actual topic says a lot, what makes my piss boil is knobs like you who are clearly Intel schills that populate online forums with bullshit about how godlike Intel is, they got caught they overstepped like they did with the 10900 that got suddenly rushed out overnight so it would be on top of a cpu bench and like the time they had a cooler onstage but claimed the cpu was running standard air, like netburst, like the HP scandal, like the antitrust suit that they lost to AMD etc etc.
They are dirty they are lowdown and frankly they make piss poor products and have for many years with Fanbois like you willing to fork out ever increasing amounts of money for a generational uplift of a few Hz being drip fed whilst at the same time gouging you with a new socklet every year remember the +, ++, +++, +++++, suddenly that wont work becasue AMD kicked ass and have been since and now Intel are back on the dirty tricks trail to keep up, as I said above whilst it was all roses Intel was prepared to let the Mobo folks do as they please but suddenly its a big issue and its all on them.
Cant wait for the real world tests to come out once all the unlocks have been locked am reading anything between 8 to 15% loss, and do you seriously expect the Mobo suppliers to just roll over and play beta, Intel just bit the hand that fed them and kicked the golden goose, retribution will be incoming for some time I reckon
They are dirty they are lowdown and frankly they make piss poor products and have for many years with Fanbois like you willing to fork out ever increasing amounts of money for a generational uplift of a few Hz being drip fed whilst at the same time gouging you with a new socklet every year remember the +, ++, +++, +++++, suddenly that wont work becasue AMD kicked ass and have been since and now Intel are back on the dirty tricks trail to keep up, as I said above whilst it was all roses Intel was prepared to let the Mobo folks do as they please but suddenly its a big issue and its all on them.
Cant wait for the real world tests to come out once all the unlocks have been locked am reading anything between 8 to 15% loss, and do you seriously expect the Mobo suppliers to just roll over and play beta, Intel just bit the hand that fed them and kicked the golden goose, retribution will be incoming for some time I reckon
TerryLaze
Polypheme
- May 9, 2016
- 9,996
- 2,695
- 63,440
Actually AMD releasing one agesa after the other for like a month proves my point and also proves that you are the shill/conspiracy nut/whatever you are.
AMD acknowledged it as an issue but you just can't see AMD doing anything bad, because I'm the fan boy and you are the level headed ranter.
https://www.tomshardware.com/news/a...t-on-ryzen-burnout-issues-limits-soc-voltages
Also the whole point is that mobo makers use "an overclocking setting" as you put it.
Yeah, stating facts makes me a fanboy, lol. Why don't you actually address what I just posted? Reviewers HAVE to choose their power limits when they boot into the bios to enable XMP. They specifically choose unlimited. They have even admitted it themselves. So what difference would that make? Are you saying Intel should have locked unlocked K cpus and not give you the option to remove power limit or what am I missing?
Ah, I was wrong. You are totally not a fanboy. You sound like a totally normal person, lol. Pointless to even talk to you, have a nice day sir.
I don't see anything bad with what AMD or Intel did. Unlocked CPUs are...unlocked. Yes , if I just turn XMP on an insanely rated kit, chances are my CPU will fry due to excessive IMC voltages (basically what happened with 7800x 3d). But that has been the case for many, many, many years, as far back as skylake in 2015. Some 4800 mhz DDR4 kits needed 1.65 (!!!) SA voltages to boot. I won't blame Intel or AMD for that, im supposed to know what the heck im doing, that's why I buy an unlocked part on an unlocked motherboard.
Same thing applies with power limits. If I remove power, thermal, and every other protection the CPU has, unless I know what the hell im doing, the CPU will fry itself. For me, that falls under "expected behavior". I don't get why people are surprised, lol. Look at me, im running prime95 at 450 watts and my CPU just died, must be AMD's / Intels fault
Last edited:
-Fran-
Glorious
- May 3, 2007
- 9,267
- 2,661
- 44,640
Well, for what is worth, they agree with blaming Intel (and by extension AMD) first and MB makers second:
View: https://www.youtube.com/watch?v=OdF5erDRO-c
Regards.
Regards.
Since this is an optional fix for Intel's worst performing CPUs, should AMD chips be retested at their base clocks and min ram spec as well? That is also a worst case scenario as AMD does not have any guarantee on their boost clocks at all, they are explicitly "up to".
Intel's baseline voltage curve on my pc with all other settings Asus "stock":
I haven't seen any reviews or heard of anyone that needed that much voltage for stability, but it is possible they exist.
Retesting could be done adhering to Intel's stated PL1/PL2 if the tests weren't performed at that standard initially, as that seems more fair, and more stable. I've got a feeling that they were though, and the unlimited power profiles given by motherboards that are stable, and used by most users actually give more performance.
Edit: I stand corrected on Tom's use of Intel PL1/PL2 - they went with motherboard values on that in their 14900k review so there may be some performance loss with that.
But another setting they have on that review is the fTPM is disabled on both AMD and Intel setups. This is a W11 requirement and if I disable my fTPM I can't use my pin and might not have a license until I reenable it. W10 doesn't need it, but W10 performs poorly with hybrid cores and is EOL next year. Having fTPM off only favors AMD and Windows 11 does not recommend turning it off unlike how they do recommend turning off virtualization for Intel CPUs. Disabling fTPM is not a setting a typical AMD user would do since it messes with their Windows 11 license and disabling it paints an unrepresentative picture of AMD's gaming performance. Maybe the games should be retested with fTPM on and in compliance with Windows 11s requirements.
In the end reviews should portray what performance a typical purchaser should expect, not the worst or best possible performance, and certainly not the worst possible performance for one brand and sweet spot performance for the other. That would be biased.
Last edited:
I've never seen such an angry face on that guy. Looks like manufactured clickbait. Might not be but it sure looks like it.
- Dec 18, 2023
- 531
- 428
- 760
The point I was making is that on non-K SKUs you can't even write a higher power limit than the fused limit into a control MSR -- you can only limit down from fused power limit, not up.
So if Intel wanted to prevent mainboard manufacturers from overriding upper limits they could have done so but they didn't because they wanted to have "teh best performance watts be damned".
- Dec 18, 2023
- 531
- 428
- 760
More like "manufactured rage", that's what most stuff on the Internet is nowadays.
vanadiel007
Distinguished
- Oct 21, 2015
- 381
- 376
- 19,060
This is what happens when you want the performance crown and beat your competition, without ensuring you are doing that within acceptable CPU limits.
What you see here are nothing more than unstable factory overclocks "approved" by Intel specs but clearly not stable and should never have been approved by Intel in the first place.
What you see here are nothing more than unstable factory overclocks "approved" by Intel specs but clearly not stable and should never have been approved by Intel in the first place.
My 13700k had the same problems. It's not just I9. I had to go into the motherboard and change settings to get power and voltages under control.
But I did that one year after trying everything else to fix the problem.
I hope after a year of use the CPU is not damaged. That would not be cool.
But I did that one year after trying everything else to fix the problem.
I hope after a year of use the CPU is not damaged. That would not be cool.
Pierce2623
Commendable
- Dec 3, 2023
- 503
- 386
- 1,260
Well they could simply do what AMD does and say that the stock settings can’t be outside the official spec. There’s a very big reason they don’t though. Review benchmarks would be a bloodbath if Intel lost 5-10% off of every score.
Pierce2623
Commendable
- Dec 3, 2023
- 503
- 386
- 1,260
The simple truth is Intel encouraged the motherboard makers to do this because it raises benchmark scores.
Were the factory overclocks on the X3D series that popped them apart also AMD approved?
It seems like Intel and AMD generally let things slide and draw hard lines that can't be crossed only sometimes.
An example would be Intel and their locked SKUs and motherboards. Another lesser example would be AMD and their 95c limit. As long as no problems arise there are no problems. Now problems have arisen for some that buy overclocking CPUs and overclocking motherboards that pre-overclock. With the answer for the instability apparently being: dial back the overclock if it is unstable.
Some Intel motherboard default settings should be changed like having the default power limit at 253w because not having that is giving some people issues. But if it isn't giving people issues or costing the CPU company money by invalidating market segmentation then it should be up to the motherboard manufacturer.
The way motherboard manufacturers do things are different. I like that. I like the freedom of choice. Asus is my favorite even though they are not perfect. Others might like different manufacturers more. Fine with me.
Also this fud over chip degradation seems just like the "Zen 2 can't handle 1.35v" hoax but with less evidence. More likely thermal paste degradation, or AIO degradation if anything. You can make an i9 put out a lot of heat.
And if you want to see over 1.6v on a 13900k, just set the motherboard to Intel's failsafe SVID and set the LLC high. It will do that with everything else stock. I don't recommend anyone use Intel's failsafe voltages unless you want to intentionally tank performance to make the chip look bad and make a video about it like HWU. Even if your chip is one of the 1% bottom bins you will still get better clocks by reducing your clocks than trying to run 1.5v all core to get 5.5GHz. No normal cooler on an unaltered CPU can handle that.
thestryker
Splendid
- Apr 19, 2016
- 4,417
- 2,667
- 30,690
I'm not going to quote anyone here, because a lot of people posting here think this is some sort of conspiracy on Intel's part when it isn't. I mentioned previously when MCE/power setting shenanigans started and Jarred mentioned how far back messing with stock settings goes. What's going on is nothing new, and just like how AMD had to change their guidance when CPUs got burned up Intel will likely have to change theirs. CPU manufacturers tend to give motherboard manufacturers a lot of leeway in various places and this just seems like too much.
Also let's face it if in fact the default settings for these boards were Intel stock none of the manufacturers would be adding Intel baseline profiles because they'd already be there.
There are a lot of Z series motherboards where the default power profile is unlimited. Many of the boards even resetting the CMOS will leave you with unlimited power. In the Z790 roundup HUB did Steve specifically mentioned this behavior:
https://www.techspot.com/review/2776-intel-z790-motherboards/
Also let's face it if in fact the default settings for these boards were Intel stock none of the manufacturers would be adding Intel baseline profiles because they'd already be there.
I mean you're clearly trolling, but I just want to point out that this part is in fact false. Intel does not stamp approval on every motherboard/BIOS.
TerryLaze
Polypheme
- May 9, 2016
- 9,996
- 2,695
- 63,440
Not really because they would also drop from 400+ watts down to 280W
You have to do a very restrictive benchmark with only two or three things to make the 14900k look bad.
https://www.techpowerup.com/review/...ke-tested-at-power-limits-down-to-35-w/2.html
But there is no default power profile. What does default even mean when you are enabling XMP? In order to enable XMP you HAVE to set your power limits first, you can't proceed unless you choose one of the 3 available options (150 // 280 // unlimited).
TRENDING THREADS
-
News TSMC founder says Intel should focus on AI, not advanced process technologies- Started by Admin
- Replies: 27
-
-
Discussion What's your favourite video game you've been playing?- Started by amdfangirl
- Replies: 4K
-
News Intel CEO Pat Gelsinger retires, effective immediately — also steps down from BOD, two co-CEOs step in
- Started by Admin
- Replies: 124
Facebook
Twitter
Instagram