Intel breaks silence on 13th and 14th-gen Raptor Lake desktop CPU instability issues

Will a software patch really be able to fix it without compromising on the quality?
I'm asking this because Intel are letting owners return their chips for replacement. Which makes me think that it's a hardware issue doomed for failure.
If AMD can come up with a witty way to get into this with an exchange ad troll it's going to be hilarious. Those "x3D" models will do wonders.
 
Will a software patch really be able to fix it without compromising on the quality?
I'm asking this because Intel are letting owners return their chips for replacement. Which makes me think that it's a hardware issue doomed for failure.
Since the damage is already done, it's only a matter of time before everyone encounters instability, so the only real fix is to provide a replacement. The micro code patch might only prevent brand new processors from getting damaged.
If AMD can come up with a witty way to get into this with an exchange ad troll it's going to be hilarious. Those "x3D" models will do wonders.
If only AM5 supported DDR4, they would already have been market leaders. DDR5 still costs ~2x of DDR4
 
Will a software patch really be able to fix it without compromising on the quality?
After going through the statement provided, it's not the processor which is faulty, but the microcode.
> "...the elevated operating voltage is stemming from a microcode algorithm resulting in incorrect voltage requests to the processor..."
So, fixing issue in microcode will resolve the issue.
 
After going through the statement provided, it's not the processor which is faulty, but the microcode.
> "...the elevated operating voltage is stemming from a microcode algorithm resulting in incorrect voltage requests to the processor..."
So, fixing issue in microcode will resolve the issue.
If a mere alteration of code will help, why do they want owners to send back their chips?
 
Does not look good for Intel at all...

Have linked a specific timestamp relevant to my comment but I would recommend watching the whole thing if you want to get a better understanding of what exactly is going on. For those who don't know, Buildzoid/AHOC is known for his in-depth guides and analyses of computer hardware. I really appreciated his (and Hardware Unboxed's) reviews of B450 motherboards back when I was building a new system in 2019, since they were one of the few places, at least back then, which tested as VRM, power delivery and VDROOP performance for as many consumer motherboards as possible, including budget ones and shared the results in an easily understandable manner.

Buildzoid also has excellent RAM overclocking guides, and would also highly recommending looking those up if you're remotely interested in extracting more performance from your memory! I personally was able to overclock 4 sticks of 3200MHz DDR4 RAM to 3933MHz with almost the same timings, leading to noticeable performance gains.
 
Last edited:

...a patch is coming in mid-August that should address the “root cause” of exposure to elevated voltage. But if your 13th or 14th Gen Intel Core processor is already crashing, that patch apparently won’t fix it.
If you have a 13th/14th gen that's running without issues, lower power and Vcore limits manually immediately, and wait until the patch drops.

Avoid buying 13th/14th gen on the used market, they might already be unstable. Don't buy brand new either since 15th Gen and AMD 9000 is on the horizon.

Sidenote: a decade ago you had to jump through hoops to OC a CPU. Now we're doing it to underclock it. How the turntables have...
 
Here is the official guidance from their forums.

June update - guidance on Intel default settings. These would have been picked up by the OEMs by now and updates should be already available. If not then they need to be manually applied.


July update - where they talk about the upcoming microcode fix


On related news AMD's 9000 lineup has been delayed to Aug as they had to pull the first batch due to QC issues.
 
Last edited:
Back
Top