Users complain on issues coz they been running on extreme profile without knowing thanks to mobo unlimited power setup for Ksku, baseline setting was only a recent update which will help fresh or new builds but for those that already have problems there is no saving, see my gearbox analogy.

If you want to know what is CPU baseline supposed to be, this is what





Anything more its no longer baseline, if you dont restrict to these limits its no longer baseline. The problem happens when mobo makers break them.

This post has been edited by babylon52281: Jul 18 2024, 01:59 AM

yes the intel base settings, like around 1 min of turbo boost then scale down. Even then why is it even if running it nonstop for 250w or 200w give problems

in the last until 11 and 12 gen, like the 8700k can be run near 200w for years without issues

That means you dont understand the crux of the issue. Running at up to max PBP didnt give problems, it is when mobos push it to 300W and above which is causing the failure.

8700k TDP is max 95W. Going 200W as you claimed is NOT NORMAL OPERATION, doesnt matter if it lasted for years. That was a fluke that nothing happened. Neither does it means it can continue to run 200W without eventual failure.

When power, thermal, voltage limits are respected, most dont have a problem.

Here is where even mobos dont respect Intel limit power profile

https://youtu.be/7Gborxw9Q_M?t=628

Still higher than 253W

This post has been edited by babylon52281: Jul 18 2024, 07:57 PM

bro people had been experience totally standard run without oc but have crashes like 3 times a day on new cpus now

https://www.tomshardware.com/pc-components/...shes-in-testing

Intel's laptop CPUs are also suffering from crashing issues — several laptops have suffered similar failures in testing

Did you not read what I wrote? Kskus will auto boost and maintain due to mobo settings. If want total reliability, get nonK. Look at business workstations, all are using nonK or vpro CPUS.

https://www.dell.com/en-my/shop/pcs-desktop...cision-desktops

https://www.hp.com/my-en/shop/desktops/busi...s-desktops.html

https://www.lenovo.com/my/en/c/workstations...Megamenu_DT_2_3

Not sure why its so difficult to understand. Peeps think Ksku is the default CPU. Nope. NonK is meant to be standard model for 90% users, Kskus are for OCers.

look at skyline post above yours

So? One example doesnt make a statistic. And I did read the Reddit post. Doesnt say anything about which laptops or which CPU was on that are failing. Very very very general jobber could say, particularly on Reddit. And there were many that whack that comment for the same thing that I brought up. So unless you have anything else factual to refute what I wrote, anecdotal comment is meaningless to me.

ok so continue on ignoring the bad experiences many other people have on their computers just like what amd fanbois did to me when I was talking about amd ryzen instability too last time. ignoring it doesn't mean it doesn't happen.

TristanX
https://www.techspot.com/news/103843-intel-...ure-limits.html

We talked about Intels efficiency & performance and it looks like Arrowlake will not be the "IT" to take back efficiency crown from AM5. Looking at the news above, it appears Intel will continue to ramp higher the thermal limit to compete with AM5 IPC gains.

As I said before, its high time Intel needs to replace Cove core with a ground up redesign aimed for a massive IPC boost. As seen by the latest issue, thermal wall limit has ady been reached and breached killing CPUS slowly.

https://youtu.be/6omKMiloLtE?t=1300

Timestamp the video this was mentioned. Mobo makers continue to ignore Intel settings even with latest default limit set bios. When paired with 13/14 Ksku it still defaults to power unlimited. Intel should really warn these mobo brands.

PSA: For owners of 13/14gen Kskus paired with whatever Z690 or Z790, in whatever bios, go in and MANUALLY SET THE LIMIT TO INTEL DEFAULT PBP 253W! Stop trusting the mobo makers will actually do their job!

This post has been edited by babylon52281: Jul 20 2024, 11:39 PM

No info on the process node. Maybe Intel wants to make it miles faster than X3D... Like Sandy Bridge vs Phenom II.

P core only CPU. Worthy upgrade from my i7-9700.
https://ark.intel.com/content/www/us/en/ark...cts/238781.html


This post has been edited by Skylinestar: Jul 21 2024, 10:06 AM

Traditionally on new node Intel would claw back the power/thermal ramp if they know it would give them an IPC edge over AMD. Coz they could then mature the node process and ramp up the power to compete. The prediction is that Lunarlake would go back to 100oC just enforces this idea. But for Arrowlake to start the power ramp game does tell me Intel will continue maintain the high power limit even after a node switch.

AM6 showed good IPC gains per gen, with 9000 Xseries almost on par with 7000 X3D, so if Intel is to target 7000 X3D is just going to lose to 9000 X3D later on. This while keeping nearly same power/thermal limit is impressive compared to what Intel been doing with 13/14 Gen and it appears Arrowlake as well.

Quote "Intel Intros 14th Gen Core "E" Embedded Processors with E-cores Disabled" Keyword: Embedded.

Also it only solve the hybrid 'Thread Missdirection' issue on Win10. If your running Win11 there isnt much to gain or might even lose some FPS vs i9 with Ecores.

This sounds similar to the upcoming Bartlett CPU gen https://www.techpowerup.com/324571/intel-pl...cessor-for-2025 but imagine a 12 PCore i9
I predict would be a thread performance monster vs AMD flagship 16cores CPUs (or even Arrowlake 8PCores i9) as it wont have latency loss running all core workloads.

This post has been edited by babylon52281: Jul 21 2024, 10:24 AM

Mlies faster = at least 30% more than 9000X3D. Means probably 60% or more than 9950X. They will go this path if they want to make sure AMD has no counter. We gotta pay for it though. Its not unexpected if Intel gets back the performance crown.

This post has been edited by TristanX: Jul 21 2024, 12:06 PM

don't think arrowlake will do 60% better than 9950X.

zen5 is amazing for achieving what it has on tsmc n4p node. while arrowlake is on tsmc n3b/intel20a node.

This post has been edited by ipohps3: Jul 21 2024, 12:11 PM

Raptor Lake = 10nm
Intel 20A = 2nm

Current AMD CPU is TSMC 4nm. Imagine Intel packing at least 400% more transistors if they use 20A, jumping from 10nm. That is a big stepup. The challenge using 20A is the heat when they pump the clockspeed.

technically raptor lake is intel7 (10nmESF) equivalent to tsmc 7nm. and intel20A is equivalent to tsmc 3nm. so about 2 generation behind only.

and arrowlake desktop i9 will be on n3b not 20A.

This post has been edited by ipohps3: Jul 21 2024, 12:53 PM

Nope.
https://www.digitaltrends.com/computing/int...ything-we-know/

Then again, its all rumors. Just wait for the final product.