CPU Thermal wall? So much for the GHz race.

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

cytg111

Lifer
Mar 17, 2008
26,810
16,078
136
I understand every word that you're saying, IDC, but I just disagree with the usage of "IPC" here.

- And that is fair, but do realize that you are the one driving and thinking "omg everyone else is drinving in the wrong lane!"
 

sefsefsefsef

Senior member
Jun 21, 2007
218
1
71
That’s exactly what im saying, CPI(IPC) has nothing to do with TIME.
...
Could you please explain what that “IPC” you are talking about really is ?? It seems to me like CPU time execution but with the Main Memory Latency included.

OK, this is making more sense now. IPC has a specific dictionary definition, which you seem to not go by. The dictionary definition is the one I'm using. Again I'll quote Wikipedia:

In computer architecture, instructions per clock (instruction per cycle or IPC) is one aspect of a processor's performance: the average number of instructions executed for each clock cycle.
...
The number of instructions executed per clock is not a constant for a given processor; it depends on how the particular software being run interacts with the processor, and indeed the entire machine, particularly the memory hierarchy.

The memory hierarchy (including the latency of the memory hierarchy) does impact IPC. The clock of the CPU ticks by, just like a clock on the wall, regardless of whether the CPU is performing an integer add, floating point divide, waiting on an L3 cache hit, waiting for a DRAM access, or waiting on disk access. It's all the same. All of these impact IPC, because they affect the number of instructions that can be completed during a certain number of clock ticks (Instructions / Cycles = IPC). Hope this helps!