Motorola MPC7457 PDF

[wizard]

Hi Gyro;

For the IMac / EMAC or even the Ibook this would have been an excellent upgrade a few months ago, now it is a matter of ho hum. On the PowerBooks it is to little to lat.

While I don't expect the PowerBook to match the G5 Power Mac I do espect it to come close. !.3 GHz is a little far from the 970 based machines for the price charged on these units.

Thanks
Dave

Originally posted by Gyroscope
C'mon people i really don't see what's so bad about this update.

This new G4 could actually be ok for PowerBook and i,eMac updates.

Although Motorola lists 1.3 Ghz as a max speed there is also another line of G4's that is sold to Apple (XC model numbers) and these are usually clocked higher. Thats why Apple was selling 1.45 Ghz PM while Motorola web site would lists max clock speeds around 1GHz. I'm positive that G4 XC could go as high as 1.6 Ghz. Of course i dont know what amont of heat this cpu would generate and if it would be suitable for PB,but it would go nicely into new consumer boxes ala iMac. Doubling amount of L2 cache is also an improvement and it really speeds up things particulary on G4/G3 architecture. It could be said that bus speed is the weakest link in G4 family of cpu's and this cpu doesn't do much to change it. Yeah measly 33 mhz increase won't do miracles but it can be offset by larger amount of L3 cache as someone previously has mentioned here. There's also that infamous Apple system controler with its few tricks to speed up things little further.

All in all seems like a decent update until IBM comes up with lower power (cooler) G5's.

 

[avus]

Correct me if I'm wrong, but these "new" G4's do NOT support true DDR.

I already knew that the 7447/7457 would not support DDR but only the "proposal" 7457-RM would, from this roadmap that was posted at MacObserver 10 months ago.

http://www.macobserver.com/images/news/2002/20021022moto/roadmap.jpg

I haven't heard about this "RM" varient of the 7457 since this report, and there are two chances that Motorola would deliever this chip - slim and none.

 

[Puppies]

What's the big deal? Isn't this what we were expecting? We already knew the 7457 is just a minor bump. And how cares? It's still a pretty decent upgrade.

Combine a 40% faster chip with a 200-400% faster GPU, and you've got quite a nice upgrade by any standards.

Yeah, Motorolla has become totally pathetic as a chip designer...but that's probably just because they can't afford it anymore. We're MASSIVLY behind where the G4 was supposed to be. I mean we were s upposed to get something massivly better than the 7457 TWO YEARS ago. But even still, it's a solid chip, and I'm still drooling over the new PowerBooks.

Now where the heck are they?

EDIT: As for these "20 watt toaster" comments. Excuse me? A G4 drawing 18 watts is still a far cry from a Pentium 4 M which can draw as much as 75 watts.

The G3 is the second best mobile chip on the market from a power/performance ratio standpoint-yes, it's better than the Pentium-M. But come one, do any of you actually want a 750FX over a 7457?

 

[AstroPlain]

Correct me if I'm wrong, but these "new" G4's do NOT support true DDR.

Hmm...maybe Motorola does suck THAT MUCH! You would think that throwing some $$ and a few years at them would lead to something good, but I fail to see it. Last good thing? AltiVec, 1998. And that put off IBM (probably the only reason for any innovation at Somerset) for 4 years.

Still, having this in a PowerBook should be a good thing, and the 7457 will surely be in the iMacs right away. Maybe as fast as 1.5 GHz, if Apple can squeeze some performance out of them.

 

[Aciddan]

Originally posted by Puppies
EDIT: As for these "20 watt toaster" comments. Excuse me? A G4 drawing 18 watts is still a far cry from a Pentium 4 M which can draw as much as 75 watts.

This is what I don't get:- we all talk about how hot the G5 is, but it's still a darn site less than what is thrown into PC laptops (they even throw in 3GHz desktop chips in some of the power models for petes sake!)

I'd just go 'bugger it' and hold off on releasing new powerbooks until a G5 could be put in one - so what if it produces more heat than a G3 or G4 - these are POWERbooks - it would still not be as toasty as a wintel lappy...

edit: I should qualify the last statement with saying - using current G5s and waiting until September.

G4: move on, nothing to see here. I thought we were waiting for a prosumer update...

-- Dan

 

[DrGonzo]

What I want to know is how some of you can say these 7457's will actually be CHEAPER when the reported yield of these chips is only at a 20% success rate. 80% seems like a LOT of wasted silicon. Maybe that other 80% can be used in lower clocked chips, but still...

 

[cuneglasus]

I'm not sure what the fuss is about,this is a minor change to a pdf document that I downloaded a couple of months ago.As usual some are jumping at some wrong conclusions.First the notion that the 7457 would support a 200 fsb is not a rumor it is clearly stated on this document several times.

http://e-www.motorola.com/files/32bit/doc/fact_sheet/PPCSALESFACT.pdf

Remember this info is mainly for their embedded customers.Apple was useing a 167 fsb 7455 for quite a while before any online document said anything about it.So they could certainly use a 200 fsb if they chose.Not that there would be a huge difference in performance.I see a lot of silly ideas about bus speeds on these forums,sepecially in regards to DDr and its performance benefits.

Similarly the 1.3 ghz maximum speed is misleading.The documents only cover the L and N versions while apple uses the P versions with higher core voltages.The 7455 low voltage cores topped out at around 1 ghz while apple and several upgrade card venders were useing parts clocked at over 40% higher (one upgrade card can do 1467 mhz).And no these are not overclocked for heavens sake (now there was a rumor/whopper)!So parts at the 1.7 to 1.8 range are not unlikely,the problem is since apple wont be useing them in the towers anymore they are not likely to go with P versions,but will use the lower voltage cores for the powerbook.This is good news for the upgrade card makers though.Hopefully some 7457 upgrades will apear very soon.

 

[raschild]

I keep getting happier and happier that Apple has started to part with Moto. Their development of the G4 has all but stopped.

 

[tdewey]

Well

Whether I buy a new 15 or 15.4" pb assuming it has this new chip in it depends more on the GPU than the CPU speed. I would buy a 1.0Ghz or 1.3Ghz Al PB if it has the new Radeon Mobility 9600 or 9600Pro in it. There are more things than the CPU these days.

On the other hand, if Apple ships a G5 pb in say Feb 2004 with an old GPU I wouldnt buy that.

 

[vniow]

Originally posted by cuneglasus
As usual some are jumping at some wrong conclusions.First the notion that the 7457 would support a 200 fsb is not a rumor it is clearly stated on this document several times.

That PDF also says:

(7457 shipping date) 7457-1.3 GHz (Feb. 2003).

and

(Typical power consumption) 7.5W @ 1.0GHz (est)

When the shipping date is actually in Q3 2003 and the new updated PDF lists the typical power consumption at 1.0ghz is:

15.8W

 

[cuneglasus]

Originally posted by vniow
That PDF also says:

and



When the shipping date is actually in Q3 2003 and the new updated PDF lists the typical power consumption at 1.0ghz is:

And the point being that a 200 mhz bus was considered and could still make its way into a shipping 7457.We just dont know.Remember its not much of a deal.A 200 mhz bus would have the same architecture as the 167,it would just support a higher frequency.

I have to wonder about those power useages.On the 7455 pdf it lists the typical useage of a 1 ghz part as 15 watts,lower than the 7457 at the same frequency.Even with a larger level two cache this seems wrong as the 7457 is made on a smaller process.The figures must be based on different conditions or something.
The 1.3 part is listed at 18.7 watts which is only about 18% or so higher for a 30% frequency increase.This one would be a good bet for the powerbook.

 

[ddtlm]

wizard:

Yeah I saw those power numbers too. Those are way higher than what I've seen before, I'm dissapointed. If those numbers are right than Moto is behind the Pentium M on watts per clock cycle, not to mention the number of clock cycles. Eek! That said, I think I'll still get some 7457's for my Quicksilver when they appear on the market.

Puppies:

As for these "20 watt toaster" comments. Excuse me? A G4 drawing 18 watts is still a far cry from a Pentium 4 M which can draw as much as 75 watts.

Thats a power figure from a desktop P4, not a P4-M, let alone a P-M.

The G3 is the second best mobile chip on the market from a power/performance ratio standpoint-yes, it's better than the Pentium-M.

How do you figure? The P-M's use/dissipate no more than 7W at 1ghz. (See page 72 of this PDF: http://www.intel.com/design/mobile/datashts/252612.htm?iid=ipp_dlc_procpmp+info_datasheet& .) The G3 is low power for sure, and not a bad processor at all, but I would bet on the P-M being faster per clock, aided by the 400mhz FSB and 1MB L2.

 

[ddtlm]

cuneglasus:

The 1.3 part is listed at 18.7 watts which is only about 18% or so higher for a 30% frequency increase.This one would be a good bet for the powerbook.

Dunno how Moto got those figures, but "true" peak power usage must scale at least as fast as clock speed (assuming nothing really strange was done with the voltage). Of course there are "untrue" types of peak power usage trickery, such as what Intel does with the P-M, where they will throttle the processor to keep it within boundaries they define, but I strongly suspect Moto is not doing that since its (apparently) pretty tricky.

 

[beloit08]

This is just sad. How can Apple continue to tout their PowerBooks as the wonderful pro machines they're supposed to be when Moto keeps pumping out junk like this? Sure, the speed bump will be nice. Sure, a slightly larger FSB will be welcome. But this is seriously getting into a situation where Moto's become the speed regulator on Apple's rental car of innovation. I think it might be time to move on and let Moto die a quiet, sad little death.

 

[excalibur313]

I don't understand why people are getting so mad about this. I'm very happy that apple is releasing a powerbook with up to a 40% boost in actual speed over the 1 ghz. For too long people have lived by the myth that going from a 1 ghz to a 2 ghz will double the speed of programs on your computer. In most cases the difference isn't as much as one would think and I think that a chip which boosts performace 40% is pretty big news. I know that no amount of money would cause me to go to a pc and I know that most of you in this forum feel the same way. Who cares if it is twice as fast if it crashes and has all of the other problems associated with windows to go along with it? Macs are classy, run fast and even a 1.3 would run almost everything as fast as you'd want. I realize that probably the people complaining are those using maya and such programs but if you really need lots of speed even a 200mhz bus won't speed it up enough to really matter in the long run and if thats the case go buy a dual 2ghz g5. That will smoke the pants off of any laptop

 

[cuneglasus]

Originally posted by ddtlm
cuneglasus:


Dunno how Moto got those figures, but "true" peak power usage must scale at least as fast as clock speed (assuming nothing really strange was done with the voltage). Of course there are "untrue" types of peak power usage trickery, such as what Intel does with the P-M, where they will throttle the processor to keep it within boundaries they define, but I strongly suspect Moto is not doing that since its (apparently) pretty tricky.

The real problem is probably that the figure they give for the 1 ghz part is wrong for some reason-too high. Dont place a lot a faith on things you see on the web,even if it is an "official" document.Mistakes are easy to make.

 

[Apple //e]

Originally posted by excalibur313
I don't understand why people are getting so mad about this. I'm very happy that apple is releasing a powerbook with up to a 40% boost in actual speed over the 1 ghz. For too long people have lived by the myth that going from a 1 ghz to a 2 ghz will double the speed of programs on your computer. In most cases the difference isn't as much as one would think and I think that a chip which boosts performace 40% is pretty big news. I know that no amount of money would cause me to go to a pc and I know that most of you in this forum feel the same way. Who cares if it is twice as fast if it crashes and has all of the other problems associated with windows to go along with it? Macs are classy, run fast and even a 1.3 would run almost everything as fast as you'd want. I realize that probably the people complaining are those using maya and such programs but if you really need lots of speed even a 200mhz bus won't speed it up enough to really matter in the long run and if thats the case go buy a dual 2ghz g5. That will smoke the pants off of any laptop

you are apple marketing´s dream come true.

the current g4s are still waaaay faster than the pentium 4 arent they? just imagine a 40% increase. wow

 

[spice weasel]

Wow, it only took Moto the better part of a year to come out with a chip that should have been put into the powerbook line back in January. Not too bad for a company that has recently fallen off the top-ten list.

The 1.3 GHz version aside, all Moto did was increase the L2 and L3 cache. How can they take so long to release a new chip, and come out with clock speeds of 867 MHz and 1 GHz -- we already have those now. Adding/doubling cache is not a real breakthrough.

The question I have is, will Apple again stick the low end 867 MHz in the 12" powerbook again? Yes double the L2 and the addition of L3 will be nice, but come on now. But despite my major dissapointment at reading the 7457 specs, I'd gladly order me up a new 12"-er if Apple would just release the new powerbooks already!!! Hell, the TiBook should have been remodeled back in January, and the 12" should have never shipped with a measly 128 megs of on-board RAM and no L3 (not to mention a crappy L2).

Moto, stick the two low-end 7457's in a cell phone or something, and give us some real breakthroughs. But I guess we can't really expect more from them, can we? After all, their slogan is "If it's over budget, behind schedule, and below expectations, it's gotta be Motorola." Apple should have dumped them back when they screwed up on the first generation of G4s.

 

[inkswamp]

Originally posted by e-coli
WOW!!!!!!!

A brand new slow-as-hell chip with outdated technology and miserable bus speeds!!!!!!!!!

Cool... just the thing to run 9.2.2!

 

[DrGonzo]

Originally posted by excalibur313
I realize that probably the people complaining are those using maya and such programs but if you really need lots of speed even a 200mhz bus won't speed it up enough to really matter in the long run and if thats the case go buy a dual 2ghz g5. That will smoke the pants off of any laptop

Of course any modern desktop will "smoke the pants off of any laptop" because it's a desktop. Most people are "complaining" about the fact that the new updates to the powerbook line aren't really updates at this point, they're more like a really stale interim stage until the g5 or whatever new fast chip comes out. Until we see such afformentioned 40% speed improvements over the current 15" tibook i think most of this community remains quite skeptical (if they do infact release a 7457 g4 in the powerbooks). YOu seem so certain that they'll definately release a 7457 in the powerbooks as well. Apple //e said it best that you indeed are a marketing dream come true to them.

 

[Phinius]

7457-RM probably on smaller process size in 2nd half of 2004

Originally posted by avus
I already knew that the 7447/7457 would not support DDR but only the "proposal" 7457-RM would, from this roadmap that was posted at MacObserver 10 months ago.

Judging by the speed improvements over the 7457, the 7457-RM seems to have been proposed for a smaller process size. Motorola has announced that a dual-cpu G4 on a chip will be made. So, that would probably be a dual-cpu G4 chip that can run as high as 2.4 GHz in the second half of 2004. Double data rate buss speeds and onboard controller are also in the works for the 7457-RM.

I haven't heard about this "RM" varient of the 7457 since this report, and there are two chances that Motorola would deliever this chip - slim and none.

The odds are very good that Motorola will produce this chip in the second half of 2004. Motorola is now a partner with two other chip manufacturers, so the rate of improvements through process shrinks should pickup considerably.

 

[Phil Of Mac]

Re: 7457-RM probably on smaller process size in 2nd half of 2004

Originally posted by Phinius


Judging by the speed improvements over the 7457, the 7457-RM seems to have been proposed for a smaller process size. Motorola has announced that a dual-cpu G4 on a chip will be made. So, that would probably be a dual-cpu G4 chip that can run as high as 2.4 GHz in the second half of 2004. Double data rate buss speeds and onboard controller are also in the works for the 7457-RM.

The odds are very good that Motorola will produce this chip in the second half of 2004. Motorola is now a partner with two other chip manufacturers, so the rate of improvements through process shrinks should pickup considerably. [/B]

Who will be buying a G4 in the second half of 2004?

That's right.

 

[Analog Kid]

Far as I can tell, the news here is this:

Mot hit .13µm
L2 cache doubled

Not sure what else people expected-- it's a G4. Still a G4 even after a shrink... Maybe DDR could have been expected, but ya'll must've thought the core would stay pretty much the same.

I'm relatively new to these boards-- did people do this much whining about the G3 once the G4 came out?

I think 1.3GHz is the fastest to expect in this spin of PB. The overclocking is done in the desktops because the power change is disproportionate to the performance change.

The L3 chache hasn't really changed. The 4MB mode where half is cache and half is "private memory" is targeted at the embedded folk.

Where's the power going, that's my question? A 20% power drop seems like less than I'd have expected-- especially since we're comparing the "pushed" version of the 7455 with the low end of the 7457 line... Is it all going into the extra L2 cache?

Guess that would make sense if the transistor counts someone posted are accurate: 60% increase in transistor count and most of it must be the cache (nothing else really changed)...

Aside from having slipped their release, it seems like about the right part to be next in line.

Not sure where the 40% performance boost is going to come from-- it's not 30% clock and 10% cache as was suggested... Performance doesn't scale with clock like that. Might be 10% clock and 30% cache...

If that's accurate it matches the expected curve-- doubling every 18 months. We're 9 months since the last update and are looking at a 40% boost. 'Nother 9 months and another 40% would hit the typical marketing curve. 9 months after that might be time for a discontinuity.


No point in comparing this against a 3GHz P4 laptop-- I don't see the point in a laptop I can't put in my lap, can't fit in my backpack, and have to plug in.

To those that keep repeating that Mot has dropped off the "top 10"-- that's a ranking of semiconductor sales. They're still ahead of IBM on that table so it clearly doesn't serve the point you're trying to make.

(I also call BS on the 20% yield numbers...)

 

[macphoria]

Looks like next PowerBook revision will be:

12" PowerBook 1GHz G4
15" & 17" PowerBook 1.3 GHz G4

 

[Analog Kid]

Re: Re: laughable

Originally posted by insidedanshead
perfectly put. no offense but this is just getting ridiculous.. we should boycott pbooks to get apple and IBMs arse on the move to get it over to a G5.. a while back I remember someone saying from apple that their relationship with Moto is still strong.. BOOOOOO.. this is a democracy .. competition... no need to be buddy buddy...give moto the cold shoulder they have done NOTHING good for us in TWO years. Bleh. Makes me sick.

Uh, really dumb question here but: how are you going to encourage competition if you're single sourced ?

Much better to have two companies leap-frogging each other...