Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

IBM's 90nm PowerPC 970FX

Oh, one more thing...

The Pentium 4 Extreme has a 2mb L3 cache running at 800mhz.

The 970 has a 800mhz to 1Ghz external bus speed. Compare this to the P4 800mhz internal bus speed (which really runs at 200mhz with a 4x multiplier). This means that the P4 only has a 200mhz to play with to access memory, which can only be double pumped to 400mhz.

Current memory in the Dual 2.0Ghz G5 runs at the equivilent of 400mhz. However, because the 1.8Ghz and higher models have dual channel memory, this can be the equivilent to 800mhz total memory bandwidth.

Of course, this means that the 970 has a L3 cache up to 8GB running at the "same" speed as the P4 Extreme 2MB cache.
 
Originally posted by jderman
Whats really important here is that now we know for sure that a g5 pb *can* be produced and that it would be a practical machine.
Click to expand...

Yes, now we know that the heat produced by the processor is not anymore as big an issue as it seemed to be just some time ago. This does *not* prove that g5 pb can be produced, there are other juice-hungry components on the G5 powermac that has to be replaced or improved before we can expect a g5 laptop according to apple standards (battery life, size, elegance).

In my opinion the most worrying components are the those that control powermac's hypertransport buses. Sure one does not need two processors in a laptop so that takes away some of the complexity but I still fear that it's gonna take some work and thus time to make them powerbook savvy.

edit: added the *not* that was missing...
 
The CPU isn't the only thing that draws a lot of power. A big system chip and fast RAM draws considerable power also. I recall Apple said the system chip in the G5 desktop was a big power drain also.

So it's nice the CPU size and power requirements have come down but that's only a part of the equation.
 
New PowerMacs, sure; iMacs, maybe; PowerBooks, no way.

This IBM document basically tells us that the 970FX can do the same work as the original 970 with half the power. Surelly this would make a re-design of the PowerMac case possible. If two more 3.5" hard drives can be squeezed in there then a lot of current G5 owners would see that as a good reason to upgrade.

While I would not go so far as to predict G5 powerbooks in the next revision, I see no reason now why iMacs could not use these new processors. From reports on the net Motorola is once again dragging its feet manufacturing G4s and surelly this will affect Apple's current momentum.
 
Originally posted by nighthawk
The 970 has a 800mhz to 1Ghz external bus speed. Compare this to the P4 800mhz internal bus speed (which really runs at 200mhz with a 4x multiplier). This means that the P4 only has a 200mhz to play with to access memory, which can only be double pumped to 400mhz.
Click to expand...

Nope. The current Pentium 4's access memory at 800Mhz using four samples in each cycle at 200Mhz. That is why the 865 and 875 use dual channel DDR400 memory. The dual channels given them an effective 800Mhz to match the processor.

Of course, this means that the 970 has a L3 cache up to 8GB running at the "same" speed as the P4 Extreme 2MB cache.
Click to expand...

Again this is incorrect. L3 cache always has much lower latency than main memory. That is one of the reasons why the G4 with 2MB of L3 cache but a pitifully slow main memory access was able to keep up with the Pentium 4's for many tasks.
 
Please clarify.....

Is the 90nm or 65nm (nano-meter?) size going to be used in the next PowerBook? I'd imagine it's going to be the 65nm, but someone please clarify this for me, I'm a little miffed on the subject.....

In other words, clue me and all the other hapless know-nothing's around here.....:)
 
Re: New PowerMacs, sure; iMacs, maybe; PowerBooks, no way.

Originally posted by Sol
This IBM document basically tells us that the 970FX can do the same work as the original 970 with half the power. Surelly this would make a re-design of the PowerMac case possible. If two more 3.5" hard drives can be squeezed in there then a lot of current G5 owners would see that as a good reason to upgrade....
Click to expand...

I agree with your thinking, I would like to add, for general purposes having two optical slots might be a higher priority(for me at least), now if they coupled that with an extra 2 drives, THEN they could get some g5 owners to shell out annother $3000 +
 
if they can get 2 2 Ghz 970fx's in a 1U server from factor, they can do it with a Powerbook too. it's coming folks. I would expect it at the same time they bring the 3 Ghz desktops.

You'll have to wait another 6 to 7 months, grasshopper. :D
 
Re: Please clarify.....

Originally posted by sethypoo
Is the 90nm or 65nm (nano-meter?) size going to be used in the next PowerBook? I'd imagine it's going to be the 65nm, but someone please clarify this for me, I'm a little miffed on the subject.....

In other words, clue me and all the other hapless know-nothing's around here.....:)
Click to expand...

I think it's highly unlikely that Apple would be able to wait for, say, a 980 (or even a 970) at 65 nM. A quick zoom through the boards here or somewhere like arstechnica will show that most believe 65nM to be some way off yet.

The PB's would be far too slow and old if they had to stick to MOT's 33 Mhz bumps til then. Sales would be nonexistent. I think you can take it to the bank that there will be PB G5s at 90nM, using the 970 fx.
 
Originally posted by nighthawk
The data-sheet for the P4e does describe the two ALU (integer calcs) as running at "twice the core clock" which is 1600mhz for the P4e. This is also unlike the 970xx which runs at the full clock speed.
Click to expand...

I don't know anything about the P4EE but the ALUs will be the same as the P4's:

By "core clock" they mean the processor speed, e.g. 3.2 GHz. The two ALUs are effectively running at 6.4GHz, making them 3 times faster than the 970's.
 
Originally posted by nighthawk
Originally posted by csubear

I am not disputing anything you say, but I just wanted to add something in the mix here...

The Pentium is only a 32-bit chip, so when the 970 is running in full 64-bit mode, it only has the equivelent of twice the cache. I agree that the L1 cache size could be improved.

Click to expand...

1. The 970 is always running in 64-bit mode when running OS X. The various OS X frameworks are just not 64-bit safe yet, so you can't safely pass 64-bit data-types to the various OS X frameworks just yet. But, you can use the 64-bit operations within your own code.

2. All 970 instructions are actually uniformly 32-bits long, even the instructions acting on 64-bit operands. The physical size of the instruction descriptor has nothing to do with its bitness. That is one of the advantages of the RISC architecture, perfect instruction alignment. Even the new mulit-operation, multi data (IBM calls "SIMOMD") are 32-bits long though they execute on 64-bit data types. So there is the same L1 Address space taken up, no more.

3. Just because its a 64-bit processor does not mean that the programmer is forced to define all operands as 64-bits long. It means that 64-bit data types are handled a lot more efficiently. So this is really a none issue when it comes to cache usage. Actually a 32-bit processor probably uses up a lot more resources when trying to do a double precision (64-bit) floating point instruction.

4. The PowerMac architecture makes the entire RAM space equivalent in speed to an L3 cache!

This is a radically better design than any other P4 or G4 system before it.
 
Originally posted by 123
I don't know anything about the P4EE but the ALUs will be the same as the P4's:

By "core clock" they mean the processor speed, e.g. 3.2 GHz. The two ALUs are effectively running at 6.4GHz, making them 3 times faster than the 970's.
Click to expand...

The P4's Integer units are double clocked but that is out of necessity. They have do do a heck of a lot of work and need it just to get by. That is why when you look at SpecInt from a per MHz point of view the 970 is still faster. It is too complicated to go into here, but as an oversimplification, each P4 Integer instruction gets broken down into multiple core instructions, whereas the 970's instructions are almost all 1 to 1. The P4 Integer unit also wears more hats than the 970's which has other units that run in parallel for some of the tasks that the P4 has has to handle in the Integer unit.

The P4 has to do a lot of work for every instruction due to its x86 legacy and CISC architecture compared to the RISC 970.
 
If Apple were AMD or Intel (talking about the P4 mobile... not the Pentium M), they'd change the packaging and market this to vendors as a laptop chip for "desktop replacement" notebooks ASAP. Unfortunately or fortunately, however you look at it, Apple has a form factor that they won't vary from and until manufacturing can be fine tuned and revised, there's no way they could put that chip in a 1 inch thick notebook. Expect our G5 PB chips to be 90nm... It's good that we're there, and I don't think Apple will wait for 65nm as its too far off, but this chip isn't ready. They'll lower the pincount likely and wait for the next stepping, at least, and hope for it to cut back 5 - 10 watts of heat.
-Kev
 
As far as W usage, all I can say is WOW. IBM has pulled off a technological miracle. According to Motorola's charts the 970FX dissipates as mush W at 2GHz as the G4 7447 at 1.3GHZ! Now we know that the G5 architecture is a lot more efficient heat and energy wise than the G4 architecture. And, the 970FX can evidently dynamically scale not only frequency but also voltage. The chart implies that the FX will dynamically scale from 1.4-2GHz and 1-1.3V.

This is really a huge advance. The only unknown is the ASIC Controller. But, we have to believe it has been refined to the 90NM process as well. This means that the 970FX will easily go into Apple's entire line of computers. This is what Steve Jobs probably meant when he said that the G5 was the MACs future at the last show. It is exactly what Apple needs to do and quickly!
 
another reason this is good for apple...

So let's throw powerbooks completely out of the equation for one second (if y'all can haha... i want one so badly).

But this is great for Apple that these chips are readily available. First of all, they are far cheaper to make. If you have a big silicon wafer and you cut it up to small chips, you can make a lot more when these chips are 90nm than when they are 130nm. While it is unlikely Apple will pass this savings onto the consumer, there's a small chance of it, but their profit margins should get higher which is better for them, as a business.

The reduced heat reduction is great news not just for laptops, but also for their desktops. I'm sure Apple really enjoyed marketing their comparmentalized air cooling systems in the G5s, and while the design is something to be proud of, the design came out of necesity, not out of anything else. The original G5s were HOT. Having all those fans in there not only increased noise, but every time you have a moving part in a PC like that based on motors and such, the failure rate goes up (more tech service bull**** for apple to deal with).

I'm wondering if we can expect a small chassis redesign from Apple with this revision. If nothing else, maybe internally. There is no need for them to have so many fans running at such a high speed. I wouldn't be surprised if they got rid of some of the fans and lowered the RPM of the remaining ones and no longer market the compartmentalized cooling system. Let's see.

-Kev
 
Originally posted by spaceballl
Unfortunately or fortunately, however you look at it, Apple has a form factor that they won't vary from and until manufacturing can be fine tuned and revised, there's no way they could put that chip in a 1 inch thick notebook.
-Kev
Click to expand...

Though I don't know what Apple will do, technically the G5 is more than capable to go into a 1" PB, even the PB 12". Re-look at those W comparisons between the G4 and the G5. These numbers tell us that Apple has a lot to work with. They can easily put a 2GHz 970FX into a 15" and 17". They can probably bump the 12" to 1.6.

More than that, the FX will easily go to 3GHz+ if you compare it to the W that the P4, even the Prescott is planning to use (120W+!). I would even venture that 4GHz is not that far away for the PowerMacs. Apple really has a loaded gun here. What a turn of events. Maybe we will see Apple surpass Intel in GHz in addition to raw performance by the end of the year.
 
Originally posted by Mr Maui
Just found a spec sheet for the 7455 at 1GHz and it says the typical consumption is 15w. Does that seem right?

Seems pretty high if the 1.4GHz G5 is only requiring 12.3w.

Any thoughts?
Click to expand...

The 7455 was produced on a 130nm process. The new G5 (970FX) is produced on a 90nm process. The smaller the circuitry, the lower the voltage and the less heat lost per cycle.

If you read the IBM pdf that this thread is about, you will see the old G5 970 on the 130nm process was as hot as the 7455 in terms of power consumption at a given speed.

I would venture the prediction that we will see G5 Powerbooks soon, with speeds according to size, maybe 1.2Ghz 12", 1.4Ghz 15", 1.5Ghz 17". It's still possible that Apple will wait for the 65nm 980 late this year and use speed-bumped G4s in the meantime, but several things make me think not: 1, At these guesstimated speeds the power consumption should be about the same in each model as the old G4 chip being replaced; 2, I don't think there are ever going to be 90nm G4s, and any faster 130nm G4s in the 15" than the ones now would reduce battery life beyond the current barely acceptable 2-3hour range; and 3, there is now poor differentiation between the iBooks and Powerbooks, and the best explanation for Apple choosing the G4 instead of the 750GX (also described in the current PDF) for the iBooks is that they are planning a rapid transition to G5 Powerbooks.

There were those people who said G5 Powerbooks are going to be impossible for some time to come because of the heat. Well, they were partially right, in that the original G5 was way too hot. But now things have changed. Apple has certainly been planning for this years in advance, and I would be surprised if they don't make a G5 Powerbook at the earliest possible opportunity.

Michael
 
Re: another reason this is good for apple...

Originally posted by spaceballl

I'm wondering if we can expect a small chassis redesign from Apple with this revision. If nothing else, maybe internally. There is no need for them to have so many fans running at such a high speed. I wouldn't be surprised if they got rid of some of the fans and lowered the RPM of the remaining ones and no longer market the compartmentalized cooling system. Let's see.

-Kev
Click to expand...

I totally agree that this will give Apple a lot of flexibility in design. What I think we will see happen to the PowerMac G5 is that the main CPU compartment will shrink. Maybe only one Fan instead of two. This will allow the G5 to have two optical drives, 4 SATA HDs and an additional PCI-X slot.

It will also allow for the G5 to enter the consumer line in a big way with smaller and radical designs. That really is what the 20th Anniversary Mac needs to be about, the consumer line which needs the most help right now.
 
Originally posted by stingerman
4. The PowerMac architecture makes the entire RAM space equivalent in speed to an L3 cache!
Click to expand...

Not true. The memory is clocked at basically the same rate at L3 cache but the latency is tremendous in comparison. L3 cache 800Mhz will always be faster than main memory at 800Mhz because of latency.
 
Originally posted by ktlx
Not true. The memory is clocked at basically the same rate at L3 cache but the latency is tremendous in comparison. L3 cache 800Mhz will always be faster than main memory at 800Mhz because of latency.
Click to expand...

True: Have you ever disabled the cache and seen how slow your computer runs?

The North Bridge is a huge bottle neck and thats why AMD has put the memory controler on the chip to remove that particular bottleneck.
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back