Intel Is Still Having Raptor Lake Crashes And Now Its Affecting Server Customers Too ▉

Wendell’s reporting here is just devastating and the follow up conversation with Steve From Gamers Nexus makes it sound like this is even worse than anyone thought.

The fact that Intel still hasn’t gotten to the bottom of this makes me think there’s a serious flaw in the architecture. Either that or there’s some serious silicon degradation happening. What makes it even worse is that it’s affecting server customers. It’s one thing when part makers like Asus or Gigabyte push the power profile or something too high and it causes an issue. We saw that when the 7800X3D launched. That’s easily fixed. If a user overclocks the CPU and causes a problem, that’s on them. Sucks, but also easily identified.

The fact that they can’t pin it down between power profiles (which they let get completely out of hand on the consumer side) and something else makes me think there’s a serious flaw in the architecture itself. The issues aren’t even consistent with the Raptor Cove (performance) cores or the Gracemont (efficiency) cores.

What’s curious is that this is also affecting server boards that are essentially just the basics and not overclocked in any way. Their power profiles are also way lower and basically at spec. The only difference is that the servers run twenty four hours a day seven days a week as opposed to end user systems that run for a period of time and then get shut down. As a result the silicon gets degraded faster than it would on some random gaming PC.

Intel better get it together with Zen 5 launching at the end of the month. They also better pray that Arrow Lake doesn’t embarrass them too.

> ▍