'Radium' wrote, in part:
| It's true that CPUs with lower clock rates can be more efficient than
| CPUs with higher clock rates if the former uses more bits-per-cycle than
| the latter.
|
| As I've recently found, the 4.7 GHz IBM Power6 holds the current record
| for highest frequency CPU:
|
| http://www.spec.org/cpu2006/results/res2007q3/cpu2006-20070708-01383.html
|
| How much does the 4.7 GHz IBM Power6 cost?
|
| Is the 4.7 GHz IBM Power6 available today? If so, where can I find it? I
| live in Southern California in Diamond Bar. Any store close to me that
| carries this processor?
_____
Do some research on your own. You are just asking a string of questions
that really have no relationship to overclocking.
To repeat, overclocking consist of swapping temperature and voltage safety
margins for higher clock speeds.
Overclocking is operating a CPU at a higher speed than that marked or
specified by the manufacturer.
| It's true that CPUs with lower clock rates can be more efficient than
| CPUs with higher clock rates if the former uses more bits-per-cycle than
| the latter.
Efficiency, what is efficiency? Efficiency can be instructions per second,
or instructions per watt, data processing accomplished per watt or second,
instructions per cycle, data processed per cycle.
| How much does the 4.7 GHz IBM Power6 cost?
|
| Is the 4.7 GHz IBM Power6 available today? If so, where can I find it? I
| live in Southern California in Diamond Bar. Any store close to me that
| carries this processor?
Nobody here knows or cares, and you can't afford it. If you had such a CPU,
what software would you run?
| What is the least expensive motherboard that is fully-compatible with
| this CPU?
See above comment.
| What is the least expensive HDD that is fully-compatible with this CPU?
Hard drives have absolutely NOTHING to do with the CPU. Modern CPUs do not
communicate directly with a hard drive. Some flavor of IDE, PATA, SATA or
SCSI interface handles data drives.
| Also, if I do buy the 4.7 GHz IBM Power6, I would like to undervolt and
| overclock it to the max possible. If I use the minimum voltage required
| for the CPU to operate, what is the highest clock frequency I can get?
| Just out of curiousity, is it physically-possible to build a CPU whose
| clock-rate is 600 PetaHz and whose Vcore is a picovolt?
The speed of light is an absolute limit to the operation of a single CPU. A
petrahertz is 1/1,000,000,000,000,000 second in duration. In that time NO
signal can travel further than 0.3 nanometer (less that the width of a
single transistor in a CPU.)
A picovolt is such an incredibly small voltage that likely someone combing
hair on the other side of the earth would change the local electrical field
by more than that.
As stated before, there is a minimum cut-off voltage for transistors - at
less than this voltage they do not "transist". This voltage is about 0.7
volts for transistors used today; additional voltage is needed to make a
clear difference between the logic 0 state and the logic 1 state, additional
voltage is needed to raise the signal above the noise level. Additional
voltage is needed to increased the switching speed. The first transistors
were built on germanium rather than silicon; these germanium transistors
have a cutoff voltage of about 0.3 volts, but have many disadvantages,
including leakage, slower switching speeds, and higher power consumption.
The short answers are NO and NO.
Phil Weldon
"Green Xenon [Radium]" <glucegen1 RemoveThis @excite.com> wrote in message
news:470a53d7$0$20609$4c368faf@roadrunner.com...
| Ed M. wrote:
| > "Green Xenon [Radium]" <glucegen1 RemoveThis @excite.com> wrote in message
| > news:47097e55$0$9578$4c368faf@roadrunner.com...
| >
| >>Phil Weldon wrote:
| >>
| >>
| >>>Overclocking is not going to damage an Intel x86 CPU before it becomes
| >>>obsolete (your Pentium 4 570J 3.8 GHz CPU already is.) On the other
| >>>hand, raising the core voltage too high can immediately destroy an
Intel
| >>>x86 CPU. Limiting the core voltage to no more than +10% should give a
| >>>good safety margin; more than that is possible (there are many accounts
| >>>of successful overclocks with a higher core voltage increase, but there
| >>>are no guarantees.
| >>
| >>Is it possible to decrease the core voltage and at the same time,
increase
| >>the clock frequency?
| >>
| >
| >
| > In almost every case, NO.
| >
| >
| >>>There is no such thing as 'extreme undervolting.' There is a cutoff
| >>>voltage at which a transistor will no longer operate. Reducing the
core
| >>>voltage will not harm a CPU, but you will rapidly reach a voltage at
| >>>which the CPU will just not operate. The CPU will not be damaged, nor
| >>>will any other part of the system. Restoring the core voltage to
normal
| >>>will set everything aright.
| >>
| >>Okay. What if I use the minimum voltage necessary?
| >>
| >
| >
| > You can try it........ That processor is pretty much set to operate on
it's
| > default Vcore and lowering it will probably make it unstable even at
it's
| > default speed.
| >
| >
| >>>Read Phil's answer carefully, and also other posts in this newsgroup.
| >>>You will benefit from a better grasp of what overclocking entails.
| >>>Basically, just trading off voltage and heat safety margins for higher
| >>>clock speeds. Slowly and methodically is the way to go.
| >>
| >>Okay.
| >>
| >>My goal here is to use the minimum voltage required and at the same
time,
| >>the highest frequency possible. Is this possible?
| >
| >
| > The best you could probably do is to find the highest possible freq
that
| > the processor will do at default Vcore, which will probably not be very
much
| > higher. What folks are trying to tell you is that the two processors you
| > mention are not the most efficient and not great OCers without voltage
| > INCREASES. They are very marginal even at default speeds and Vcore.
Those
| > processors were Intel's last go-round at using raw speed rather than
better
| > efficiency (operations per cycle) like the Core 2 Duos and Quads. An
E6600
| > at 2.4Ghz outperforms my EM64T 630 OC'd to 3.6Ghz by almost double in
any
| > benchmark. The Q6600 does even better, even in apps that are not
optimized
| > for it. Almost everyone in this group are very knowledgeable and are
really
| > trying to help you out. You just have to listen to what they are telling
| > you........Regards
| >
| >
| >
| > Ed
| >
| >
|
| It's true that CPUs with lower clock rates can be more efficient than
| CPUs with higher clock rates if the former uses more bits-per-cycle than
| the latter.
|
| As I've recently found, the 4.7 GHz IBM Power6 holds the current record
| for highest frequency CPU:
|
| http://www.spec.org/cpu2006/results/res2007q3/cpu2006-20070708-01383.html
|
| How much does the 4.7 GHz IBM Power6 cost?
|
| Is the 4.7 GHz IBM Power6 available today? If so, where can I find it? I
| live in Southern California in Diamond Bar. Any store close to me that
| carries this processor?
|
| What is the least expensive motherboard that is fully-compatible with
| this CPU?
|
| What is the least expensive HDD that is fully-compatible with this CPU?
|
| Also, if I do buy the 4.7 GHz IBM Power6, I would like to undervolt and
| overclock it to the max possible. If I use the minimum voltage required
| for the CPU to operate, what is the highest clock frequency I can get?
|
| I believe Power6 uses more instructions-per-cycle than most Intel or AMD
| processors. So it is a combination of higher-clock-rate and more
| bits-per-cycle.
|
| Just out of curiousity, is it physically-possible to build a CPU whose
| clock-rate is 600 PetaHz and whose Vcore is a picovolt? >> Stay informed about: Overclock but undervolt -- safe? How much? 
FAQ
Profile
Private Messages
Log in
