I don't have any numbers, but from my understanding they tend to run a little cooler in the case since their heat distribution is in a more confined area.
The dual 64-bit core PowerPC970MP? (970MP) is the next evolutionary step in the PowerPC 970 family of microprocessors. The higher frequency grade versions of the 970MP consume higher amounts of power than earlier IBM microprocessors do, and that can cause temperature issues. Each 970MP processor core contains a thermal diode used to monitor its operating temperature. The thermal diode must be monitored to ensure that the maximum operating temperature of the 970MP is not exceeded. These thermal diodes used by the microprocessors are unique to the PowerPC architecture and cannot be ?read? using standard thermal diode monitoring chips.
I'm kinda disappointed that there will be (ostensibly) 4-processor desktop G5 systems and not a single portable G5, though ...
Quad cores! I see my predicted "Octos" (see my sig since 2000) is about to finally be realized!
My read of the IBM PDF is that "Thermal Diodes" will be what allows the G5 Powerbook to have a 970MP without melting the keyboard. How about a "dual" Powerbook?
Of all the information presented by this most recent publication, IBM's innovative usage of thermal diodes is the most revealing.
The PowerPC 970MP microprocessor contains dual 64-bit microprocessor cores, double L2 caches (1MB each), and a Processor Interface (PI) bus arbiter operating above the Gigahertz speed range. The higher clock rate 970MP microprocessors dissipate larger amounts of power than did earlier IBM microprocessors (refer to the PowerPC 970MP Datasheet for the values).
Hmm... this seems to imply that IBM took an even simpler path than I was thinking (dual-core with shared L2). It looks like what they did was take the 970fx, double the L2, fabricate 2 of them on the same die, and glue them together with an elastic bus (PI) switch. This will be a ~170 million transistor, ~150 sq mm processor. At 2.5 GHz, probably dissipates ~90 Watts?
Hopefully they also did some circuit tuning and process tuning for power efficiency as well. I'm waiting on a Powerbook G5, or maybe even better, a cheaper Powerbook G4.
Quad core, or more I think is inevitable, but not this round. 2x Dual core PowerMac is fairly obvious, and there will probably be a 1x Dual core PowerMac as well.
Quad processor x Dual core (octo) will not happen. Apple has never made a Quad processor socket PowerMac before I don't see them starting that now.
-----------------------------
THT, do you really think they glued them together?
THT, do you really think they glued them together?
Text illustration:
Code:
--------------------------
| ---------- ---------- |
| | 1 MB L2 || 1 MB L2 | |
| |----------||----------| |
| | 970 || 970 | |
| | core || core | |
| ----..----------..---- |
| | PI bus arbiter | |
| ----------..---------- |
------------..------------
||
1+ GHz PI bus
||
to
U3 System ASIC
or
next gen northbridge
The PI is, surprisingly in an IBM document, the Apple Processor Interconnect, aka Apple PI, aka elastic bus, is it not?
So, it sounds like they took two complete 970 "GX" processors, except for a complete elastic bus arbiter, fabricated them on one die, and tied them together with an on-die elastic bus switch. L2 to L2 transfers, if done properly, can go through the switch, and not back to the northbridge. I'm taken the "double L2 caches" to mean that they are independent of each other and specific to their core.
That's the very definition of multi-core though. And likely not worse or better than a large unified, shared L2 for vast majority of cases. Still needs a large, high speed off-die backside L3 cache though.
THT, are you saying that the process used in making this new chip is a stopgap project that was put together in a pinch? In your opinion, will this be a compettive chip with AMD/Intel?
In figure 3, V core is shown driving the sense circuit.
Do I take this as a maximum core voltage of 2.348Volts?
What does that mean for core frequency?
If the V core of the 970FX is 1.3 @2.5Ghz, what are we
talking about here? Certainly not a possible max of 4.5 Ghz?
Anyone have an idea?
The voltage the 970fx receives is variable based on power-tune power management. No real relationship, directly, with CPU frequency.
Also, from the PDF that is on the page Aphelion linked to above:
The higher frequency grade versions of the 970MP consume higher amounts of power than earlier IBM microprocessors do....
They do indeed produce more heat than the 970fx currently at higher frequencies. PowerTune is about all Apple is doing as far as controlling the heat (other than the obvious radiator bolted to the current PowerMacs). I don't see this happening in a PowerBook any time soon.
It seems to me like introducing dual dual core 2.5Ghz+ G5s is a little much at this point. Seems like the prices would have to go up right? You're not going to get a 4 processor system for 2500... i mean seriously. Perhaps apple is going to create a workstation type computer out of these new chips, but I doubt the regular PowerMac will get 4 processors just yet. Seems like overkill honestly. I am sticking with the next upgrade just being a speed boost + PCIe etc.
I think it would be good if apple created a dual dual core workstation PowerMac X series or something starting at 3500 or so.
I mean I just don't know, but it def seems like overkill to me.
Comments
The dual 64-bit core PowerPC970MP? (970MP) is the next evolutionary step in the PowerPC 970 family of microprocessors. The higher frequency grade versions of the 970MP consume higher amounts of power than earlier IBM microprocessors do, and that can cause temperature issues. Each 970MP processor core contains a thermal diode used to monitor its operating temperature. The thermal diode must be monitored to ensure that the maximum operating temperature of the 970MP is not exceeded. These thermal diodes used by the microprocessors are unique to the PowerPC architecture and cannot be ?read? using standard thermal diode monitoring chips.
So it would seem it's coming to a Macintosh soon.
Originally posted by Aphelion
CONFIRMED!
So it would seem it's coming to a Macintosh soon.
Nice catch!
Originally posted by hmurchison
Nice catch!
M.Isotope over in the Mac Archia forum deserves the credit, but I've always wanted to post CONFIRMED!
Originally posted by hmurchison
Nice catch!
Indeed!
I'm kinda disappointed that there will be (ostensibly) 4-processor desktop G5 systems and not a single portable G5, though.
I may just have to break down and finally get a desktop Mac.
The text mentions the temp diode and monitoring countless times,
but there was no mention at all of on or off die memory control.
Could "MP" possibly mean Memory Processor?
I've read many posts on this, but still don't know if
Apple will use the 970MP "as is" or a derivative with on die memory
control.
I guess this means that the Power5 is out for this round.
Originally posted by audiopollution
What! No quad-cores?
I'm kinda disappointed that there will be (ostensibly) 4-processor desktop G5 systems and not a single portable G5, though ...
Quad cores! I see my predicted "Octos" (see my sig since 2000) is about to finally be realized!
My read of the IBM PDF is that "Thermal Diodes" will be what allows the G5 Powerbook to have a 970MP without melting the keyboard. How about a "dual" Powerbook?
Of all the information presented by this most recent publication, IBM's innovative usage of thermal diodes is the most revealing.
The PowerPC 970MP microprocessor contains dual 64-bit microprocessor cores, double L2 caches (1MB each), and a Processor Interface (PI) bus arbiter operating above the Gigahertz speed range. The higher clock rate 970MP microprocessors dissipate larger amounts of power than did earlier IBM microprocessors (refer to the PowerPC 970MP Datasheet for the values).
Hmm... this seems to imply that IBM took an even simpler path than I was thinking (dual-core with shared L2). It looks like what they did was take the 970fx, double the L2, fabricate 2 of them on the same die, and glue them together with an elastic bus (PI) switch. This will be a ~170 million transistor, ~150 sq mm processor. At 2.5 GHz, probably dissipates ~90 Watts?
Hopefully they also did some circuit tuning and process tuning for power efficiency as well. I'm waiting on a Powerbook G5, or maybe even better, a cheaper Powerbook G4.
Dual Core 2.5Ghz
Dual Dual Core 2.5Ghz
Quad Dual Core 2.5Ghz
?
Trying to keep up with the discussion.. Is a "Quad Dual Core" a possibility? Anyone care to guess at a range of time when we might see these?
Thanks
Quad processor x Dual core (octo) will not happen. Apple has never made a Quad processor socket PowerMac before I don't see them starting that now.
-----------------------------
THT, do you really think they glued them together?
Originally posted by onlooker
THT, do you really think they glued them together?
Text illustration:
--------------------------
| ---------- ---------- |
| | 1 MB L2 || 1 MB L2 | |
| |----------||----------| |
| | 970 || 970 | |
| | core || core | |
| ----..----------..---- |
| | PI bus arbiter | |
| ----------..---------- |
------------..------------
||
1+ GHz PI bus
||
to
U3 System ASIC
or
next gen northbridge
The PI is, surprisingly in an IBM document, the Apple Processor Interconnect, aka Apple PI, aka elastic bus, is it not?
So, it sounds like they took two complete 970 "GX" processors, except for a complete elastic bus arbiter, fabricated them on one die, and tied them together with an on-die elastic bus switch. L2 to L2 transfers, if done properly, can go through the switch, and not back to the northbridge. I'm taken the "double L2 caches" to mean that they are independent of each other and specific to their core.
That's the very definition of multi-core though. And likely not worse or better than a large unified, shared L2 for vast majority of cases. Still needs a large, high speed off-die backside L3 cache though.
[Edit: illustration...]
Voltage(mV) = (Fuse string/2) + 300
(range = 300 to 2348mV)
In figure 3, V core is shown driving the sense circuit.
Do I take this as a maximum core voltage of 2.348Volts?
What does that mean for core frequency?
If the V core of the 970FX is 1.3 @2.5Ghz, what are we
talking about here? Certainly not a possible max of 4.5 Ghz?
Anyone have an idea?
Originally posted by THT
The PI is, surprisingly in an IBM document, the Apple Processor Interconnect, aka Apple PI, aka elastic bus, is it not?
Yes, it is. IBM used the Apple PI term internally during the development of U2 and the first GPUL.
Originally posted by stuckinwintel
In figure 2 of the pdf a voltage range is given..
Voltage(mV) = (Fuse string/2) + 300
(range = 300 to 2348mV)
In figure 3, V core is shown driving the sense circuit.
Do I take this as a maximum core voltage of 2.348Volts?
What does that mean for core frequency?
If the V core of the 970FX is 1.3 @2.5Ghz, what are we
talking about here? Certainly not a possible max of 4.5 Ghz?
Anyone have an idea?
The voltage the 970fx receives is variable based on power-tune power management. No real relationship, directly, with CPU frequency.
Also, from the PDF that is on the page Aphelion linked to above:
The higher frequency grade versions of the 970MP consume higher amounts of power than earlier IBM microprocessors do....
They do indeed produce more heat than the 970fx currently at higher frequencies. PowerTune is about all Apple is doing as far as controlling the heat (other than the obvious radiator bolted to the current PowerMacs). I don't see this happening in a PowerBook any time soon.
I think it would be good if apple created a dual dual core workstation PowerMac X series or something starting at 3500 or so.
I mean I just don't know, but it def seems like overkill to me.