Your PC is as stable as the last piece of software you successfully ran. That's unfortunately the truth; no pc is 100% stable, you can't guarantee it
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
Intel has a Pretty Big Problem..
- Thread starter Purgatory
- Start date
Your PC is as stable as the last piece of software you successfully ran. That's unfortunately the truth; no pc is 100% stable, you can't guarantee it
I'm not ruling it out but every time someone comes up with "signs" of mass 13th/14th gen failures so far when the dust settles it is either a tiny number of affected systems or something else, so with a backdrop of issues with the game I'm wary as to the root cause.
Interestingly there seems to be common failure combinations which appear more regular in the game, some of it may reflect volume of systems:
-14700K shader compilation.
-13900K and 9950X start up, shader compilation and random crashes.
-9800X3D random crashes (often with ASRock motherboards - possibly people still on older BIOS).
Last edited:
Aye, possibly. It's interesting that this one seems to be tripping up the 13700K and 14700K, when it's more commonly the **900s we see going down to the degradation.
For context ASRock have the same problem Asus had with cooking the SoC on the 7800X3D, ASRock are doing it with the 9800X3D, dead chips have burn marks on the SoC voltage pads, 86% of 9800X3D's failure rates are on ASRock boards, compared to 4% and 1% on MSI and Gigabyte boards.Well i just Pikachu faced..... not.Stay away from ASRock motherboard's people...
I'm sure this isn't related, right?
Associate
- Joined
- 26 May 2008
- Posts
- 1,976
A big part of that was sheer volume of sales.
Asrock sold over 2 million AM5 motherboards, and the 9800X3D is one of the most popular, yes t despite this, and the issue affecting all 9000 series X3D, apparently just over 100 failed X3D CPUs
The majority on a bios that was advised to change, and some where inexperienced were playing with bios settings.
Add to that users blaming later bios updates after an early bios did the damage.
It puts a damper on owning an Asrock board, but many other motherboard brands have their own issues.
Hopefully one day we won't have children designing consumer products.
Whoever decided to actually check the CRC when decompressing assests is defintely off Intel's Xmas list!
Does sort of remind me of ECC as in most of assume memory errors do not exits, but those running server grade hardware and collecting kernel stats know otherwise. The rest of us silently corrupt data!
Still the worst GPU margins of the three by far. And AMD barely bother with GPUs ATM as they consider them to be poor margin per wafer - a trule flawed strategy on AMD's part IMO if we consider fixed costs. Speaking of fixed costs...
Intel's fixed costs of bringing Battlemage and Alchemist to market must be huge. And this is Intel who's attention span for anything not x86 CPUs is notoriously short - the only difference with GPUs is that CPUs without a GPU are pretty unsellable so they need to continue with iGPU. And somehow, the pressure of a dGPU is actually good for their drivers - as their previous approach of "it is stable in Windows and 5 games - no need for any more updates" is not viable in 2025+
While there is an easy to follow guide for ASRock boards, to be fair if you set the BIOS up properly to fully avoid the supposed root cause of the issue there are a few less than trivial changes which need to be made beyond just updating the BIOS and/or pitfalls to avoid if actively tweaking the system.
Intel has fantastic 3rd Qtr. Post 4Bn profit. Share price doubles.
https://www.tomshardware.com/pc-com...ith-operating-profit#xenforo-comments-3888078
https://www.tomshardware.com/pc-com...ith-operating-profit#xenforo-comments-3888078
on the other hand
Intel earnings aren’t swaying the skeptics: ‘This fight is far from over.’
“We understand the desire to claim victory for the embattled company, but this fight is far from over,” a Bernstein analyst says.
Last edited:
Soldato
- Joined
- 28 Sep 2014
- Posts
- 3,740
- Location
- Scotland
Alleged Intel Core Ultra X7 358H and Ultra 5 338H Cinbench R23 score leaks online - VideoCardz.com
Core Ultra X7 358H and Ultra 5 338H score 20K and 16K points, respectively, in Cinebench R23 A site called Laptopreview Club has already leaked 3DMark Time Spy performance figures for Intel’s next-gen Core Ultra 300 series, but it appears they have more data to share, this time involving a CPU...
videocardz.com
Ouch Panther Lake Core Ultra X7 358H at 65W on Intel 18A is over 9% SLOWER than Arrow Lake Core Ultra 7 255H at same 65W on TSMC N3B, Intel is in huge trouble.
Accorded to Intel 18A paper at VLSI 2025 back in June claimed CPUs on Intel 18A will expect to have 25% faster performance than on TSMC N2. Accorded to google TSMC N2 process have 15% faster performance than TSMC N3B so Panther Lake Core Ultra X7 358H supposed to be about 40% faster than Arrow Lake Core Ultra 7 255H on Cinebench R23 scores.
Panther Lake CPU will have underwhelm launch at CES 2026 and nobody will be interest to buy it. Intel 18A is now proved to be a massive disaster for Intel, they should cancelled 18A process and switched Panther Lake to TSMC many months ago, Intel 18A is slower and has very low yield than TSMC N3B.
Alleged Intel Core Ultra X7 358H and Ultra 5 338H Cinbench R23 score leaks online - VideoCardz.com
Core Ultra X7 358H and Ultra 5 338H score 20K and 16K points, respectively, in Cinebench R23 A site called Laptopreview Club has already leaked 3DMark Time Spy performance figures for Intel’s next-gen Core Ultra 300 series, but it appears they have more data to share, this time involving a CPU...
Ouch Panther Lake Core Ultra X7 358H at 65W on Intel 18A is over 9% SLOWER than Arrow Lake Core Ultra 7 255H at same 65W on TSMC N3B, Intel is in huge trouble.
Accorded to Intel 18A paper at VLSI 2025 back in June claimed CPUs on Intel 18A will expect to have 25% faster performance than on TSMC N2. Accorded to google TSMC N2 process have 15% faster performance than TSMC N3B so Panther Lake Core Ultra X7 358H supposed to be about 40% faster than Arrow Lake Core Ultra 7 255H on Cinebench R23 scores.
Panther Lake CPU will have underwhelm launch at CES 2026 and nobody will be interest to buy it. Intel 18A is now proved to be a massive disaster for Intel, they should cancelled 18A process and switched Panther Lake to TSMC many months ago, Intel 18A is slower and has very low yield than TSMC N3B.
Few things you conveniently forgot to mention
The 358 has 4 performance cores, the 255 has 6. Assuming it's not engineering samples with slow clocks, then yes 9% slower cpu than 255h but doing it with 50% less P cores so not bad really.
Then you say to yourself if they cut P cores by 50% they saves power so why is it 65w, where did that power go? It went to the GPU, the 358's GPU is 75% faster than the 255
Last edited:
Soldato
- Joined
- 15 Oct 2003
- Posts
- 16,341
- Location
- Chengdu
How dare you use logic and say anything positive about Intel. This isn’t allowed in 2025.
They both have 16 cores total.
Core Ultra X7 358H: 4+8+2
Core Ultra 7 255H: 6+8+4
Yes it has 4 P-Cores vs 6, that is a good efficiency improvement, if true, depending on what the actual power consumption was in that test, remember Intel likes to quote TDP below reality (these days as do AMD) for example they quote a TDP for the 265K as 125 watts and yet in Blender (similar to Cinebench) its actually pulling 160 watts.
As a comparison a Ryzen 9600X scores 17,460 at 87 watts, the Core Ultra X7 358H scores 15% higher if 20,000 points, Zen 5 is built on TSMC N4P, Zen 6 will be built on TSMC N2P.
Core Ultra X7 358H: 4+8+2
Core Ultra 7 255H: 6+8+4
Yes it has 4 P-Cores vs 6, that is a good efficiency improvement, if true, depending on what the actual power consumption was in that test, remember Intel likes to quote TDP below reality (these days as do AMD) for example they quote a TDP for the 265K as 125 watts and yet in Blender (similar to Cinebench) its actually pulling 160 watts.
As a comparison a Ryzen 9600X scores 17,460 at 87 watts, the Core Ultra X7 358H scores 15% higher if 20,000 points, Zen 5 is built on TSMC N4P, Zen 6 will be built on TSMC N2P.
Last edited:
I pat my undervolted 13600k on the head... It's OK, my little trooper, don't listen to the youtubers!
Last edited:
Also for desktop gaming, does any one really give a crap about peak wattage?
Critical for a laptop I guess for battery power, but a difference of 50w or even 100w at peak load when benching or gaming etc. is neither nor there in terms of your annual electricity bill.
Remember we are talking about peak power consumption here, during very demanding loads, not day to day usage when any cpu should be down clocking and using reletivley less power regardless.
Critical for a laptop I guess for battery power, but a difference of 50w or even 100w at peak load when benching or gaming etc. is neither nor there in terms of your annual electricity bill.
Remember we are talking about peak power consumption here, during very demanding loads, not day to day usage when any cpu should be down clocking and using reletivley less power regardless.
Soldato
- Joined
- 24 Jun 2021
- Posts
- 4,678
- Location
- Oxon
makes the room hot
I mean technically sure, but also anyone who can afford a high wattage PC can also afford air conditioning to keep the room cool