I've just finished building a new Windows machine using the i5-650 CPU. There was some doubt, when I bought the CPU, as to which version of it I'd end up with, as Intel have produced two different Steppings for it - C2 and K0. I thought it'd be marked on the CPU itself but that didn't turn out to be the case. How can I tell if I've got a C2 or a K0? I wrote down the markings on the CPU:
Intel M C '08 i5-650
Intel R Core i5
For this particular chip, what's the optimum average core voltage at which to operate it? I gather that it can be operated at anywhere between 0.65v and 1.4v. However, I'm not a gamer or anyone else who would want to overclock it and, indeed, I'm wanting to operate it so as to achieve merely the published working CPU frequency (3.20GHz) and with minimum power consumption. My System BIOS, which I've just updated, is showing that the CPU is running at 3331 MHz (3.33GHz), and the CPU core voltage is apparently 1.220v. I've put all BIOS speed settings to standard, so at present this chip is supposed to be working at the Intel default speed. But to which precise core voltage can I use my BIOS to adjust the CPU speed downward slightly, to get 3.20 GHz spot-on? My BIOS gives a whole range of voltages, from about 0.5v to 1.5v, to choose from. Is the optimum average core voltage for this chip more like 1.10v?
Have there been any confirmations of a reported issue with i5 processors used with Intel P55/H55 chipsets that involves a negative value for the core temperature? I'm using this i5 CPU on a Gigabyte motherboard and a number of people on the Gigabyte forums have described a problem where, after a period, the System BIOS gives the CPU core temperature as a highly negative value. This causes the cooling fan to all but stop. The CPU then, as you'd expect, overheats. This could be, of course, just down to a badly-written BIOS. Gigabyte tech support has not commented on it yet.
You will find all the information from the link below, select the product which you have got based on the product code.
Also refer to the respective datasheet if needed.
All the best,
Thanks for that link, Rajiv. It's answered ONE of my questions. I've found that the Stepping is C2. However, nothing in the detailed specification or elsewhere there gives any indication as to the nominal or optimum average voltage at which the core can be operated. Intel must surely have a nominal core voltage for the chip, yet they don't seem to publish it. Instead, Intel just publishes the allowable operating range, that being 0.65v to 1.40v in this case.
Is the nominal core voltage (to get the chip to run at precisely 3.20 GHz) exactly halfway between 0.65v and 1.40v, for instance? Will, in fact, the nominal voltage vary from sample to sample?
The previous Intel processor I used was a Pentium 4 and there was no problem in finding out what the nominal core voltage for that was, and so I can't understand why it's so incredibly difficult to find out the same thing for the i5-650.
Not sure if there is a nominal value as from my experience each individual CPU is different, even identical models one needs mor evoltage than teh other, one is hotter than the other, so that might be teh reason.
If tehre is a nominal value and intel do publish it you will find it here: http://www.intel.com/design/corei5/documentation.htm
a lot has changed since the original Pentium 4 was released captain
unless you've got an 'Extreme', 'X' or 'K' chip (we are talking about an i5-650 and not an i5-655K correct?) then those firmware voltage manipulations will most likely have no effect... if the firmware does indeed somehow succeed in directly manipulating those voltages then you'll be demonstrating the information you've already discovered which would be the maximum voltage tolerance range for this model of chip . . . although each "piece of silicon" is different so those represent ideal\extreme temporary cases for new chips on certified systems and not what each and every i5-650 ever manufactured can manage in any situation
as far as acheiving the correct clockspeed you already have!! different software calculates identical numbers differently and considering both voltage and technically even clockspeed are analog numbers being manipulated by digital mechanisms 't's nowonder the numbers can be a bit misleading
almost every bios ever written uses the same "IBM PC Compatible" former-standard of 0.5 millisecond polling interrupts (as far as I know) which may sound like a lot (or a little) but it nothing compared to the multi-nanosecond degrees of precision available today! if you view the following link Compare Intel® Products and "highlight differences" you'll the only technical diference between the i5-650 and i5-650 are the "Bus/Core Ratio" and "Clock Speed" which are actually 2 parts of the same coin...if you divide Ark's given Clock Speed of the i5-650 by its given Bus/Core Ratio you'll get 0.1333 GHz (3 repeating) [3.2 GHz / 24 tocks = 0.133] or 133 MHz...the "back side bus" or 'bClk' almost every i686 and x86-64 processor has used forever
the fact your bios reports 3.33 GHz means either
- there's a bug that's adding an additional tock into the calculation
- you actually own an i5-660 not i5-650
- you own an i5-655k and your bios/chipset is automatically overclocking one tock
- your processor is turboboosting for some odd reason
- some other issue "above my paygrade"
hope this helps!