Intel Xeon MTBF/Lifetime

I'm looking for hard numbers involving the Mean Time Between Failure (MTBF) or lifetime that can be expected of the Intel Xeon 5000-series of processors. I'm just looking for stock data (no overclocking, or running at above-average temperatures, or so forth).

I've looked through Intel's official sources (i.e. the processor datasheets) and haven't found any relevant information. I'm wondering if anyone here knows of a source (hopefully reputable) for the data.

Thanks.

I can't say that I've seen MTBF numbers on any CPU as you would on, say, storage hardware.

Nor can I say that I've ever seen a CPU failure (other than dead-on-arrival chips) that I didn't suspect was caused by external forces.

Heat, power issues, and (most often) human handling kill CPUs much more often than CPUs die on their own; that's the nature of it when they have no moving parts or parts that degrade over time.

In lieu of concrete numbers (if anyone can provide some, I'll be very interested), I'll provide an estimate: years longer than you'll be using it.

I can't give you solid numbers, but I can offer this anecdote:

I have never seen a CPU fail of its own accord. I have seem them fail because of failed cooling devices, I have seen them fail because of a power surge on the motherboard. I have seen dozens (hundreds?) of motherboards fail. I occasionally see some RAM arrive DOA.

But I have never seen a CPU fail on its own. Not even the really crappy Cyrix 686 processors of the Socket 7 days.

Case in point: Last week I upgraded the processor in an old Watchguard II firewall from an AMD K6-2 200 to a K6-2 500. The processor I used for the upgrade was about 12 years old, and the processor I replaced has been in service for almost 10 years.