Pentium 4 'performance throttling' debunked


— 12:00 AM on April 16, 2001

Kyle at the HardOCP has done a nice job summarizing his thoughts on the recent ruckus over the Pentium 4's temperature-based clock throttling feature. This analysis by Bert McComas of InQuest Market Research points out that the Pentium 4 has a built-in temperature monitor that effectively slows the chip's clock rate if temperatures rise beyond a certain point. The analysis claims this feature saps Pentium 4 performance in performance-critical applications.

Trouble is, none of the plethora of benchmarks I've run on the P4 has shown the kind of unpredictable, uneven performance one might expect if heat-senstive clock throttling had come into play. Kyle's experience has obviously been much the same. He writes:

We have been running an over-volted overclocked Pentium 4 with the factory heatsink installed now for some time. It has been running here beside my desk folding proteins for Stanford University now for a solid month and has stayed at 100% CPU utilization. I track its performance and I can assure you that it has not ever slipped into the throttling that Bert speaks about above.
Kyle backs up his claim with a statement from Intel that confirms observations: "You can run benchmarks all day . . . unaffected by the thermal protection circuitry."

Not only that, but having fried a few T-birds himself, Kyle notes that such a thermal protection feature might not be a bad thing at all.

I'm inclined to agree. Throwing up a cloud of FUD about the Pentium 4's performance in very, very, very extreme conditions isn't helpful. Perhaps Bert will be able to do some testing and show P4 performance dropping off severely in a relatively normal thermal environment. But I suspect he may have to revise his views on this matter, instead.

 
   
Register
Tip: You can use the A/Z keys to walk threads.
View options

This discussion is now closed.