Soldato
I am actually on AM4, and am talking purely about performance. I am not taking anything into account like power envelope or (almost certainly not as high as 50%) failure rates.
Plus these things are recent, not over the last decade.
-
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
I am actually on AM4, and am talking purely about performance. I am not taking anything into account like power envelope or (almost certainly not as high as 50%) failure rates.
Plus these things are recent, not over the last decade.
It’s hard to say what these failures will look like, as by this point in time these parts should be well into most stable part of the bell curve with minimal premature failure rates. I think over time this issue could become worse.
We don't really know what the ultimate failure rate is yet aside from not being zero - it could be anywhere from point something of a percent through to 100%. There are a fair few forum posts from people with 13th or 14th gen i9 issues but only a subset have details which largely confirm as being this problem specifically and those reports in reality are a tiny number compared to the number of i9 CPUs sold - though that doesn't mean other people aren't suffering from crashes and either living with it or simply replacing or returning the CPU, etc. and a tiny number of individual cases of i7s which may be this issue. Then there is a mix of companies from suppliers to end users with large setups reporting anything from very low to very high failure rates, which could be explained as being a batch related problem maybe, or a difference in use case being more or less likely to expose the issue and so on.
Last edited:
We can deduce with some certainty that these failures are going to become worse over time. Intels failure bell curve seems upside down.
To a degree but it still isn't known exactly what is going on and whether limited to certain batches or not. One SI for example is saying they've not (so far) seen degradation with systems built near release but a certain number of those were faulty out the box and the replacement CPUs are the ones which have been degrading.
Associate
This issue is spanning three generations and all parts seem effected to some degree. Power use (essentially frequency) seems to have a statistically significant impact and is being used as a mitigation strategy to prolong life span or regain stability but seems a temporary fix. Considering the power use of these chips some operators may have already deployed this mitigation to reduce opex.
I’d say Intel have more than a single issue to resolve plus all the pending issues. If any of what we’re dropping another node clanger then it’s going to be a big problem for the whole industry.
Soldato
It's OCUK tradition though
