Install the app
How to install the app on iOS
Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
Interesting video discussing issues with 13th and 14th Gen Intel CPUs
- Thread starter wreeder
- Start date
That's an interesting video. I have two Dell computers and neither suffer from crashes like that. My desktop has a 13900K and my new Alienware M18 has the 14900K. I've used Trainz and other programs on both for hours at a time without seeing anything with my desktop running for 16 to 18 hours day while putting in 6 to 8 hours a day in Trainz and other graphics and CPU intensive programs. I also do not overclock either system even though I am capable of doing so. Being a hardware tech who went through an engineering program, I'm all too familiar with component tolerances and always operate my hardware within the limits its capable of rather than push to the highest and degrade the components.
With that said, it makes me wonder if there are more than a couple of things going on. Intel may have an obvious QC issue with the latest batches of chips, or are maybe digging through the bins and selling CPUs from previous batches to clear up old inventory. I'm thinking the latter. Intel built up a huge inventory of chips during COVID-19 due to people not being able to afford to build systems and they need to clear up what stock is sitting their warehouse before building more.
From what I recall about CPU manufacturing, the chips have a step-value. This step-value is the equivalent of a manufacturing revision process. The lower the step-value, the earlier the batch. As time goes on, the later step-values offer the best stability due to the tweaks and refinements in the manufacturing process. Since these processor lines are new, then the step-values are still pretty low, and the chips are not as stable at higher clock speeds as people imagine. This may also explain the quirky performance even on the stable motherboards where the datacenters found lowering the CPU clock and memory speeds helped the stability.
The other issue as he said is motherboard and system manufacturers. ASUS, MSI, and many others really do push the systems and then the user goes in and tweaks the systems more. We know that Intel allows overclocking but we also know that users will push their systems right to the bleeding edge of stability so that the slightest resonant spike in a coil will cause the system to crash. Combine this with the lower step-level chips and it's a recipe for disaster.
Everything may be running at the edge to start with and then the user tweaks things a bit more and the end result is these push the hardware more than it's meant to run at.
With that said, it makes me wonder if there are more than a couple of things going on. Intel may have an obvious QC issue with the latest batches of chips, or are maybe digging through the bins and selling CPUs from previous batches to clear up old inventory. I'm thinking the latter. Intel built up a huge inventory of chips during COVID-19 due to people not being able to afford to build systems and they need to clear up what stock is sitting their warehouse before building more.
From what I recall about CPU manufacturing, the chips have a step-value. This step-value is the equivalent of a manufacturing revision process. The lower the step-value, the earlier the batch. As time goes on, the later step-values offer the best stability due to the tweaks and refinements in the manufacturing process. Since these processor lines are new, then the step-values are still pretty low, and the chips are not as stable at higher clock speeds as people imagine. This may also explain the quirky performance even on the stable motherboards where the datacenters found lowering the CPU clock and memory speeds helped the stability.
The other issue as he said is motherboard and system manufacturers. ASUS, MSI, and many others really do push the systems and then the user goes in and tweaks the systems more. We know that Intel allows overclocking but we also know that users will push their systems right to the bleeding edge of stability so that the slightest resonant spike in a coil will cause the system to crash. Combine this with the lower step-level chips and it's a recipe for disaster.
Everything may be running at the edge to start with and then the user tweaks things a bit more and the end result is these push the hardware more than it's meant to run at.
PortLineParker
UK Route Builder
I have the 14900k and have had for some months now and I've not run into any issues whatsoever, it is hands down the best CPU I have ever owned and I would recommend to anyone. I do use monitoring software and when running Trainz the hottest it tends to get is around the 60 degree mark which is a far cry from my old CPU, the 11700k, which used to reach 90 degrees easily. I do have a CPU AIO cooler but I find it was worth the money as I now have a powerful system that doesn't run hot when gaming, and as we all know Trainz can be quite resource heavy.
Cheers,
PLP
Cheers,
PLP
Last edited:
I should look into that model. I have an i9-9900K with an NVidia RTX 3080Ti, and the CPU almost always bottlenecks my GPU with heavier games (Trainz included). I don’t overclock the CPU, and I’d rather get an upgrade than push the one I have. The problem is my motherboard. It’s a Gigabyte Aourus Z390 Ultra that can only take 9900 series CPUs. I would have to buy a new motherboard to go with a new CPU and my case. What motherboard is in your machine?I have the 14700k and have had for some months now and I've not run into any issues whatsoever, it is hands down the best CPU I have ever owned and I would recommend to anyone. I do use monitoring software and when running Trainz the hottest it tends to get is around the 60 degree mark which is a far cry from my old CPU, the 11700k, which used to reach 90 degrees easily. I do have a CPU AIO cooler but I find it was worth the money as I now have a powerful system that doesn't run hot when gaming, and as we all know Trainz can be quite resource heavy.
My thoughts.
That's a large number of CPUs. I was one of the lucky ones and wasn't affected by the issue with my two Dell computers.Intel has admitted there are faulty units out there and are offering free replacement CPUs. The number seems to be about 25% of chips are effected.
I wonder how the exchange will work for consumers who purchased from companies such as Dell or HP for instance? Are they going to get a chip in the mail, or are they SOL unless they bring it up?
With laptops, this means returning the unit back to the manufacturer to exchange the motherboard for another since a surface mounted CPU is difficult to replace due to the number of pins and pads.
PortLineParker
UK Route Builder
I should look into that model. I have an i9-9900K with an NVidia RTX 3080Ti, and the CPU almost always bottlenecks my GPU with heavier games (Trainz included). I don’t overclock the CPU, and I’d rather get an upgrade than push the one I have. The problem is my motherboard. It’s a Gigabyte Aourus Z390 Ultra that can only take 9900 series CPUs. I would have to buy a new motherboard to go with a new CPU and my case. What motherboard is in your machine?
My thoughts.
My 11700k was on a Z590 Aorus Elite AX but the 14900k is on a Z790 Aorus Master, running alongside a 4080-S.
Cheers,
PLP
livingsteam
Active member
Where did you see this?Intel has admitted there are faulty units out there and are offering free replacement CPUs. The number seems to be about 25% of chips are effected.
Intel is releasing a firmware update to fix the problem. I can't find anywhere that they are replacing CPUs.
It might be like the Pentium exchange in the 90s. Intel hired techs to swap out the cpus for customers that received notice that their chips had the issue. I can't remember what the issue was but the zoo had 4 machines that were effected. I took them to a local hotel where the techs had set up shop in a conference room and they swapped them out in about an hour. I guess Intel could foot the bill for manufacturers to replace whole units.That's a large number of CPUs. I was one of the lucky ones and wasn't affected by the issue with my two Dell computers.
I wonder how the exchange will work for consumers who purchased from companies such as Dell or HP for instance? Are they going to get a chip in the mail, or are they SOL unless they bring it up?
With laptops, this means returning the unit back to the manufacturer to exchange the motherboard for another since a surface mounted CPU is difficult to replace due to the number of pins and pads.
The firmware update fixes processors that aren't damaged. If the processor was overclocked and became damaged, this firmware update isn't going to fix anything. Just a good idea not to tweak the settings to the point of getting too hot and self-destruction ensues.Where did you see this?
Intel is releasing a firmware update to fix the problem. I can't find anywhere that they are replacing CPUs.
It was a letter sent to a datacenter from Intel. Of course, it could have been targeted to just datacenters but that is certainly the customers screaming the loudest at the moment. It is one thing if an end user has this problem versus a customer using thousands of these chips. If I understand the issue correctly the firmware fix just stops the cpu from accepting too much power from the MB if it is available. That is not something a CTO would accept when performance effects the bottom line.Where did you see this?
Intel is releasing a firmware update to fix the problem. I can't find anywhere that they are replacing CPUs.
I remember that. Where I worked at the time, Intel sent a rep in and he did the swap out in a conference room on all the machines we brought in for him. He was there for quite a long time due to the floor to ceiling stacks of machines he had to deal with.It might be like the Pentium exchange in the 90s. Intel hired techs to swap out the cpus for customers that received notice that their chips had the issue. I can't remember what the issue was but the zoo had 4 machines that were effected. I took them to a local hotel where the techs had set up shop in a conference room and they swapped them out in about an hour. I guess Intel could foot the bill for manufacturers to replace whole units.
This brings me back to my early point. Some of the motherboard manufacturers push the spec right to the max to begin with then the overclocking nerds go further. If the chips are marginal to begin with, then this puts them right on the edge of failure with everything else pushing them right over.The firmware update fixes processors that aren't damaged. If the processor was overclocked and became damaged, this firmware update isn't going to fix anything. Just a good idea not to tweak the settings to the point of getting too hot and self-destruction ensues.
From what I remember, the fix is changing the default optimized BIOS profile to not be 'wide-open throttle' for unlocked processors. After it is applied, folks can still adjust it as to what they want. The difference is going to be the default optimized profile is not going to allow the excessive power through it. It is a BIOS update, not really a fix inside the processor. But I could be wrong. I didn't try to study it much since I went with 12th generation and not unlocked. It is not really all that much better in reality to leave it overclocked while doing day to day computer tasks or gaming. You can always just adjust the power profile of your system in Windows Control Panel to performance or ultimate performance if you want to be safer and still get the most out of the system. For Linux, I could not explain the equivalent, if there is one.It was a letter sent to a datacenter from Intel. Of course, it could have been targeted to just datacenters but that is certainly the customers screaming the loudest at the moment. It is one thing if an end user has this problem versus a customer using thousands of these chips. If I understand the issue correctly the firmware fix just stops the CPU from accepting too much power from the MB if it is available. That is not something a CTO would accept when performance effects the bottom line.
livingsteam
Active member
It was a letter sent to a datacenter from Intel. Of course, it could have been targeted to just datacenters but that is certainly the customers screaming the loudest at the moment.
@wreeder
My Google skills don't seem to be up to snuff, I have tried several searches and I can't find any Intel letter to datacenters talking about replacing chips. You'd think it would be easy to find LOL.
And maybe I could see that happening for some preferred customers, but if the only people with damaged/unfixable chips are those who overclocked them, I seriously doubt Intel is going to make them whole.
If you have a link to that letter please post thanks!
I just hope that my freezes (although fortunately are rare) when shutting down TRS22 aren't related to this (although the previous build that I had freezed more when after editing and visiting the test track it was unable to shut down properly). I've bought a 13600K in February and as far as I'm aware, I didn't suffered any problem related with the CPU. The most worrisome thing was when after downloading some content for Trainz, VIDs spiked to 1.56 volts, but nothing else. When I'm in game temperatures barely exceed 50ºC, although VID is between 1.28-1.3something volts.
Well 13th / 14th generation Intel CPUs are being described as 'ticking timebombs' in the tech press. I'm hearing the Intel update is distinctly a short term bodge that doesn't actually fix the problem long term. So I'm now considering my first ever AMD system built around an R7 7800X3D. Intel have brought this upon themselves!
Paul
Paul
Similar threads
- Replies
- 3
- Views
- 758
- Replies
- 1
- Views
- 697