Li, Aubrey wrote:
> Mark.Haywood at Sun.COM wrote:
> 
>> Mark Haywood wrote:
>>> Li, Aubrey wrote:
>>>> Mark.Haywood at Sun.COM wrote:
>>>>
>>>>> Li, Aubrey wrote:
>>>>>> Alexander Kolbasov wrote:
>>>>>>
>>>>>>> Attached is the version that prints
>>>>>>>
>>>>>>> Status of virtual processor 0 as of: 12/05/2007 12:01:46
>>>>>>>  on-line since 11/26/2007 15:06:47.
>>>>>>>  The i386 processor operates at [1596, 2128, 2660] 2660 MHz,
>>>>>>>        and has an i387 compatible floating point processor.
>>>>>>> ...
>>>>>> This one looks good to me.
>>>>> I should point out that I've seen at least one platform (MacBook
>>>>> Pro I think) that exported about 10 P-states.
>>>> So are you suggesting we should display a range like [1596 MHz -
>>>> 2660 MHz]?
>>> I do feel strongly about either format, but thought others might if
>>> they knew it was possible (but probably rare) to see 10 frequencies
>>> listed. 
> For some users don't know speedstep, a range like [1596 MHz - 2660 Mhz]
>  will let them think the processor can change the speed smoothly.
> If the range display is prefered, I suggest to give the number of how
> many steps out.
> 
>> Make that, "I do not ..."
> So, what's your thoughts? Keep the current one? But I think Alex's idea
> is good, ;-)

If I thought it was common to see more than 4-5 P-states I'd prefer the 
range. My minimal experience has been that more than 4 is uncommon. So, 
if I absolutely had to give an opinion? Yeah, I think that the separate 
list of frequencies is better.

Mark

Reply via email to