> "Sorry, unable to believe: 2-4% failure rate for CPU's?"
It seems high to me too. I can't recall ever having a CPU fail and I must have used/owned hundreds of them in my lifetime. But presumably, failure rates in data centres where CPUs are run 24/7 at high temperatures etc are higher than in consumer applications?
Interesting that failure rates seem to peak in the summer months, too, and this didn't seem to be explained in the article. Perhaps the data center's cooling is working less effectively in summer?
It seems high to me too. I can't recall ever having a CPU fail and I must have used/owned hundreds of them in my lifetime. But presumably, failure rates in data centres where CPUs are run 24/7 at high temperatures etc are higher than in consumer applications?
Interesting that failure rates seem to peak in the summer months, too, and this didn't seem to be explained in the article. Perhaps the data center's cooling is working less effectively in summer?