RapidIO will tie together Motorola's PowerQuicc, C-Port and PowerPC G4 families, he said.
Looks like the G4 family will get RapidIO.
Rooster
G4 is dead for us (I hope)
Quote:
Quote from article
"Motorola supplies a lot of silicon to Cisco. One of the advantages of PowerQuicc III is that we can integrate other IP [intellectual property] fairly quickly. If Cisco says they need a certain interconnect capability or some other specific technology, we have the capability to integrate that."
Too bad those Butt-heads wouldn't do a damn thing for the Mac community when we needed something.
ehhh? I think you missed the point of the quote. It basically dispels the rumor that Moto cancelled the 7457. As of April 3 of this year, Moto's plan is to extend the G4 line (among others) to include RapidIO (or interestingly enough, Hypertransport). This is the 7457. Therefore Moto themselves is saying that they are continuing development...not quite what the original rumor said, is it? Slow to adopt or not, I hope Moto does continue to develop the G4 (and beyond). Competition is good, particularly for Apple.
Good catch, Rooster.
onlooker, doesn't it make you wonder why Moto didn't do that for Apple? And why they would for Cisco? I personally think there is a lot more to this story than meets the eye, and certainly more than anyone will ever say!
onlooker, doesn't it make you wonder why Moto didn't do that for Apple? And why they would for Cisco? I personally think there is a lot more to this story than meets the eye, and certainly more than anyone will ever say!
I think it's a merger of three things.
Moto will say anything to sound good in front of press, and public.
They can't produce serious products on their own in time to stay competitive.
They fxcked Apple for stopping the clone plant
You are right though. There is certainly more to this than anyone at motorola will ever admit to.
BTW, they would have lost their @$$ if that clone sh*t would have went through. The economy was about to do a nose dive.
looking back they probably owe Apple millions for shuting them down before it got going.
Yes, maybe someday. But since the relevant bit was one line from a very lengthy post that basically says RapidIO is seeing less industry skepticism, and, oh yes "hooray. we're soon to ship the MPC85xx", I wouldn't count on seeing in a G4 anytime soon. As in, not for at least a year from today. If they don't change the roadmap before then, that is.
onlooker, doesn't it make you wonder why Moto didn't do that for Apple? And why they would for Cisco?[/B]
They're not doing anything for Cisco. Nothing. All the Moto rep said was "if Cisco wants to buy the PowerQuic III (MPC85xx) we can add any silicon they want." Meaning, "We don't care, we'll add HT to it, as long as Cisco is willing to pay..."
But Cisco isn't going RIO, they've standardized on HT. So, no, Cisco isn't buying.
Asked if Motorola's networking group would add a HyperTransport module to PowerQuicc III specifically for Cisco Systems, which backs the HyperTransport interconnect developed by Advanced Micro Devices, Perkins said, "Motorola supplies a lot of silicon to Cisco. One of the advantages of PowerQuicc III is that we can integrate other IP [intellectual property] fairly quickly. If Cisco says they need a certain interconnect capability or some other specific technology, we have the capability to integrate that."
This says to me that the rumor about Motorola dropping support of the G5 for Apple due to Apple wanting their own interconnect Architosh (rumored to be hypertransport) is BS.
Is it possible the Motorola G5 exists, is ready to go, and is just waiting for the PPC 970 to be released? Seems to me it died quite suddenly. Maybe the G5 is for high end macs and the ppc 970 is for mid and low end?
First, the 7457 isn't going to get cancelled, and there was no rumor to that effect. The '57 is late, but it should show up this year. The rumored cancellation was of the 7457-RM, an itself rumored, RapidIO-enabled version of the G4.
The modular core is a product of the Book E spec jointly developed by IBM and Motorola. It's designed to allow bits and pieces to be added easily for customers. The G4 is not, has never been, and will probably never be made to the Book E spec. It's too powerful (stop laughing).
If the 7457-RM does appear, great. It'll offer a serious boost to G4 performance, and near-compatibility with the 970 boards (assuming that Apple goes with RapidIO as the thread in their motherboard fabric). By the time it appears the G4 should be very small, very cool and very cheap, still packing an impressive power/watt ratio and no longer nearly as memory bound. It'll be a great CPU for inexpensive Macs, iBooks, and more powerful Digital Hub devices. The AltiVec unit will be ideal for the sort of compute-intensive coding and decoding that video and audio require.
It might be an attractive chip for Apple, if Apple is still on speaking terms with Motorola. It might be less attractive if IBM comes through with their G3-that's-really-a-G4, or if a similar competitor appears. It's important to remember that whatever Mot's foibles (and they are legion), the G4 is a very nice CPU on its own terms. It's not an ideal CPU for a $3500 workstation, but it isn't meant to be.
First, the 7457 isn't going to get cancelled, and there was no rumor to that effect. The '57 is late, but it should show up this year. The rumored cancellation was of the 7457-RM, an itself rumored, RapidIO-enabled version of the G4.
See, this is where I get lost.
The G4s were the 7400, 7450, 7455, and now/soon the 7457 and 7457-RM.
The "cancelled" chip was the 7_5_00. Yes? The "G5"? The one rumored to be floating around in the "welded" boxes and "exploding"?
They've extended their 8xxx series to the 85xx... because there's a market for those. But if your major purchaser has said "No, that's crap, we aren't going to buy it", there's no reason to push the 75xx to market.
I could be confused, but the 7457 doesn't seem like a 'generational change' sort of chip. More like the way the G3 has progressed to smaller/cooler/whatnot. It's still a G3.
No conspiracy needed to explain Moto's problems. Incompetence explains it just fine, and it all boils down to the fact that Moto's HiP 7 fab hasn't shipped anything of merit yet. As should be known, the 7457, 8540 and 8560 chips are waiting on it to ramp up. The fab is going on 18 months late, and looks to be 2 years late until Moto can ship said chips in volume.
18 to 24 months is an entire fabrication technology design cycle. Being 1 year late is horrible, 18 to 24 is disastrous. Maybe by some miracle they'll get fab time on the 90 nm fab they are building with STM, Philips et al, but I doubt that fab will ramp up until mid-2004.
No conspiracy needed to explain Moto's problems. Incompetence explains it just fine, and it all boils down to the fact that Moto's HiP 7 fab hasn't shipped anything of merit yet. As should be known, the 7457, 8540 and 8560 chips are waiting on it to ramp up. The fab is going on 18 months late, and looks to be 2 years late until Moto can ship said chips in volume.
18 to 24 months is an entire fabrication technology design cycle. Being 1 year late is horrible, 18 to 24 is disastrous. Maybe by some miracle they'll get fab time on the 90 nm fab they are building with STM, Philips et al, but I doubt that fab will ramp up until mid-2004.
There, now see!
Two paragraphs really can explain just about everything.
No conspiracy needed to explain Moto's problems. Incompetence explains it just fine, ....
True or not this really made me laugh.
Just a note - that eweek article mentions nothing about the MPC 7457-RM. As far as I'm concerned its'(the MPC 7457-RM) status is still only proposed, which is the way a Motorola PDF document listed it quite some time ago.
Another thing that struck me funny is the reference to the potential "160,000-unit-per-year" "cellular-basestation market". This quote, "Motorola's flagship communications control processor. By RapidIO standards, PowerQuicc III is a big engine" makes it sound like Motorola bent over backwards for this 160,000 unit market. Yet they wouldn't take a financial risk in developing a G5 in an acceptable time frame for Apple and the potential for millions of sales.
I know I'm leaving a lot out concerning the MPC 8540/MPC8560 and PowerQuicc markets, but to me this is how the article reads.
The G4s were the 7400, 7450, 7455, and now/soon the 7457 and 7457-RM.
The "cancelled" chip was the 7_5_00. Yes? The "G5"? The one rumored to be floating around in the "welded" boxes and "exploding"?
The (again, rumored) 7500 was apparently cancelled as well. I was responding to a poster above who seemed to say that the 7457 would be cancelled as well. As THT has mentioned, it's not been cancelled, just postponed for too long. The rumors were that the memory-controller-sporting, RapidIO-enabled version of the '57 was a goner. As well as the 7500.
Quote:
I could be confused, but the 7457 doesn't seem like a 'generational change' sort of chip. More like the way the G3 has progressed to smaller/cooler/whatnot. It's still a G3.
If the 7457 is a 7455 with a die shrink and a bigger cache, and the 7457-RM is a 7457 with a few performance enhancements and an on-die memory controller with a RapidIO interface, then we're not talking about generational changes. The 7457 and the 7457-RM are just G4s. The 7500 was something else altogether, if the few scraps of rumors about it were right.
Mot won't make what they can't sell, but (again as THT points out) they also can't sell what they can't make. That's probably where everything went wrong.
Another thing that struck me funny is the reference to the potential "160,000-unit-per-year" "cellular-basestation market". This quote, "Motorola's flagship communications control processor. By RapidIO standards, PowerQuicc III is a big engine" makes it sound like Motorola bent over backwards for this 160,000 unit market. Yet they wouldn't take a financial risk in developing a G5 in an acceptable time frame for Apple and the potential for millions of sales.
That 160,000 unit market is just one target for the PowerQuicc III -- there many many many others, if they can deliver their product. The earlier PowerQuicc chips were a multi-billion dollar revenue generator for Motorola, and they're hoping the new ones will be too (if they ever get them out the door). They are probably right too because they are great chips for the markets they are aimed at.
As Amorph said, one of the key features of the PowerQuicc III design is rapid customization / modularity. Moto didn't have to bend over backwards to address a particular need in this market, they just had to lean a little to the side and none of their other customers would even notice. Developing a G5 (or G4+RIO) is a much more expensive undertaking with only one customer (at least for the G5, they might have others for a G4+RIO which is why they still mention it occasionally to see if anyone asks for it).
That 160,000 unit market is just one target for the PowerQuicc III -- there many many many others,.......
Programmer, you're absolutely correct.
That's why I said,"...makes it sound like Motorola bent over backwards for this 160,000 unit market."
I also said,"I know I'm leaving a lot out concerning the MPC 8540/MPC8560 and PowerQuicc markets, but to me this is how the article reads." I really did understand the market is much larger than 160,000.
It just struck me funny, I guess my sense of humor is somewhat warped. For years now, I've only had one gripe w/ Motorola. My gripe pertains to their seeming unwillingness in the past 3 years to take a chance on Apple and develop a cpu specifically designed for a desktop, not just make the G4 OK for a desktop The G4 has fallen, arguably, woefully behind the competition.
Oh the irony, of years of hoping Motorola would take that chance, then reading an article making it sound like Motorola bent over backwards for 160,000 made me laugh(out loud at that).
It would have behooved the author of the article to mention the total potential market for these processors not just the cellular base station market,...
And it would have behooved me to be clearer in my post.
Was MOT's problem that they didn't want to "take a chance" with Apple or was it that they couldn't get their Fabs to work to guarantee delivery to Apple. Recent info seems to suggest the latter.
I've never seen the contract between Apple and MOT so I don't know what requirements it incorporated. Apple may have pulled a non-performance clause out of the contract and said "give us this processor performance by a certain date or we're looking elsewhere".
Was MOT's problem that they didn't want to "take a chance" with Apple or was it that they couldn't get their Fabs to work to guarantee delivery to Apple. Recent info seems to suggest the latter.
Motorola have many more problems than an inability to get their new process working. They are desperately short of chip designing talent, and have real problems keeping engineers at the company. The company is notorious as a bad place to work, Dilbert's early columns were based on employee experiences at the company!
The SPS is perenially underfunded (thus the problems with HiP 7), you can't do semiconductors on the cheap. When you cut budgets your best talent often walks, into better jobs at Intel, IBM and, most frequently, AMD. If you don't have senior chip engineers around you also loose experience, your design ethos and methedologies and historical perspectives. Worse, you loose the ability to train new engineers in the practical methods of design.
Senior engineers are at the "tiller" of the design process. They make sure the general direction of the project is kept on course, deadlines are met and problems are fixed. Without that level of design experience it is extremely difficult to deliver a chip on-time, on-budget and within the target design parameters (frequency, performance, die area and power draw being the major ones).
The 7500 project was outside of all parameters, late into the design cycle with major attrition of the engineers working on it.
Motorola made the only desicion they really could. They were in no position to bet the company on this late and expensive behemoth, so it got dropped and the remaining engineering talent redistributed to other achievable projects.
Comments
I have found this at:
http://www.eetimes.com/story/OEG20030331S0019
RapidIO will tie together Motorola's PowerQuicc, C-Port and PowerPC G4 families, he said.
Looks like the G4 family will get RapidIO.
Rooster
Originally posted by Rooster
[B]Back at the begining.
I have found this at:
http://www.eetimes.com/story/OEG20030331S0019
RapidIO will tie together Motorola's PowerQuicc, C-Port and PowerPC G4 families, he said.
Looks like the G4 family will get RapidIO.
Rooster
G4 is dead for us (I hope)
Quote from article
"Motorola supplies a lot of silicon to Cisco. One of the advantages of PowerQuicc III is that we can integrate other IP [intellectual property] fairly quickly. If Cisco says they need a certain interconnect capability or some other specific technology, we have the capability to integrate that."
Too bad those Butt-heads wouldn't do a damn thing for the Mac community when we needed something.
Originally posted by Rooster
Back at the begining.
I have found this at:
http://www.eetimes.com/story/OEG20030331S0019
RapidIO will tie together Motorola's PowerQuicc, C-Port and PowerPC G4 families, he said.
Looks like the G4 family will get RapidIO.
Rooster
In 2001, GREAT
In 2004
PPC970 arrives 8)
Motorola F*** OFF
Originally posted by Fat Freddy
In 2001, GREAT
In 2004
PPC970 arrives 8)
Motorola F*** OFF
ehhh? I think you missed the point of the quote. It basically dispels the rumor that Moto cancelled the 7457. As of April 3 of this year, Moto's plan is to extend the G4 line (among others) to include RapidIO (or interestingly enough, Hypertransport). This is the 7457. Therefore Moto themselves is saying that they are continuing development...not quite what the original rumor said, is it? Slow to adopt or not, I hope Moto does continue to develop the G4 (and beyond). Competition is good, particularly for Apple.
Good catch, Rooster.
onlooker, doesn't it make you wonder why Moto didn't do that for Apple? And why they would for Cisco? I personally think there is a lot more to this story than meets the eye, and certainly more than anyone will ever say!
onlooker, doesn't it make you wonder why Moto didn't do that for Apple? And why they would for Cisco? I personally think there is a lot more to this story than meets the eye, and certainly more than anyone will ever say!
I think it's a merger of three things.
- Moto will say anything to sound good in front of press, and public.
- They can't produce serious products on their own in time to stay competitive.
- They fxcked Apple for stopping the clone plant
You are right though. There is certainly more to this than anyone at motorola will ever admit to.BTW, they would have lost their @$$ if that clone sh*t would have went through. The economy was about to do a nose dive.
looking back they probably owe Apple millions for shuting them down before it got going.
Originally posted by Rooster
Looks like the G4 family will get RapidIO.
Rooster
Yes, maybe someday. But since the relevant bit was one line from a very lengthy post that basically says RapidIO is seeing less industry skepticism, and, oh yes "hooray. we're soon to ship the MPC85xx", I wouldn't count on seeing in a G4 anytime soon. As in, not for at least a year from today. If they don't change the roadmap before then, that is.
Originally posted by Thai Moof
Good catch, Rooster.
onlooker, doesn't it make you wonder why Moto didn't do that for Apple? And why they would for Cisco?[/B]
They're not doing anything for Cisco. Nothing. All the Moto rep said was "if Cisco wants to buy the PowerQuic III (MPC85xx) we can add any silicon they want." Meaning, "We don't care, we'll add HT to it, as long as Cisco is willing to pay..."
But Cisco isn't going RIO, they've standardized on HT. So, no, Cisco isn't buying.
Asked if Motorola's networking group would add a HyperTransport module to PowerQuicc III specifically for Cisco Systems, which backs the HyperTransport interconnect developed by Advanced Micro Devices, Perkins said, "Motorola supplies a lot of silicon to Cisco. One of the advantages of PowerQuicc III is that we can integrate other IP [intellectual property] fairly quickly. If Cisco says they need a certain interconnect capability or some other specific technology, we have the capability to integrate that."
This says to me that the rumor about Motorola dropping support of the G5 for Apple due to Apple wanting their own interconnect Architosh (rumored to be hypertransport) is BS.
Is it possible the Motorola G5 exists, is ready to go, and is just waiting for the PPC 970 to be released? Seems to me it died quite suddenly. Maybe the G5 is for high end macs and the ppc 970 is for mid and low end?
First, the 7457 isn't going to get cancelled, and there was no rumor to that effect. The '57 is late, but it should show up this year. The rumored cancellation was of the 7457-RM, an itself rumored, RapidIO-enabled version of the G4.
The modular core is a product of the Book E spec jointly developed by IBM and Motorola. It's designed to allow bits and pieces to be added easily for customers. The G4 is not, has never been, and will probably never be made to the Book E spec. It's too powerful (stop laughing).
If the 7457-RM does appear, great. It'll offer a serious boost to G4 performance, and near-compatibility with the 970 boards (assuming that Apple goes with RapidIO as the thread in their motherboard fabric). By the time it appears the G4 should be very small, very cool and very cheap, still packing an impressive power/watt ratio and no longer nearly as memory bound. It'll be a great CPU for inexpensive Macs, iBooks, and more powerful Digital Hub devices. The AltiVec unit will be ideal for the sort of compute-intensive coding and decoding that video and audio require.
It might be an attractive chip for Apple, if Apple is still on speaking terms with Motorola. It might be less attractive if IBM comes through with their G3-that's-really-a-G4, or if a similar competitor appears. It's important to remember that whatever Mot's foibles (and they are legion), the G4 is a very nice CPU on its own terms. It's not an ideal CPU for a $3500 workstation, but it isn't meant to be.
Originally posted by Amorph
A few quick points here:
First, the 7457 isn't going to get cancelled, and there was no rumor to that effect. The '57 is late, but it should show up this year. The rumored cancellation was of the 7457-RM, an itself rumored, RapidIO-enabled version of the G4.
See, this is where I get lost.
The G4s were the 7400, 7450, 7455, and now/soon the 7457 and 7457-RM.
The "cancelled" chip was the 7_5_00. Yes? The "G5"? The one rumored to be floating around in the "welded" boxes and "exploding"?
They've extended their 8xxx series to the 85xx... because there's a market for those. But if your major purchaser has said "No, that's crap, we aren't going to buy it", there's no reason to push the 75xx to market.
I could be confused, but the 7457 doesn't seem like a 'generational change' sort of chip. More like the way the G3 has progressed to smaller/cooler/whatnot. It's still a G3.
18 to 24 months is an entire fabrication technology design cycle. Being 1 year late is horrible, 18 to 24 is disastrous. Maybe by some miracle they'll get fab time on the 90 nm fab they are building with STM, Philips et al, but I doubt that fab will ramp up until mid-2004.
Originally posted by THT
No conspiracy needed to explain Moto's problems. Incompetence explains it just fine, and it all boils down to the fact that Moto's HiP 7 fab hasn't shipped anything of merit yet. As should be known, the 7457, 8540 and 8560 chips are waiting on it to ramp up. The fab is going on 18 months late, and looks to be 2 years late until Moto can ship said chips in volume.
18 to 24 months is an entire fabrication technology design cycle. Being 1 year late is horrible, 18 to 24 is disastrous. Maybe by some miracle they'll get fab time on the 90 nm fab they are building with STM, Philips et al, but I doubt that fab will ramp up until mid-2004.
There, now see!
Two paragraphs really can explain just about everything.
Take the rest of the day off THT
Originally posted by THT
No conspiracy needed to explain Moto's problems. Incompetence explains it just fine, ....
True or not this really made me laugh.
Just a note - that eweek article mentions nothing about the MPC 7457-RM. As far as I'm concerned its'(the MPC 7457-RM) status is still only proposed, which is the way a Motorola PDF document listed it quite some time ago.
Another thing that struck me funny is the reference to the potential "160,000-unit-per-year" "cellular-basestation market". This quote, "Motorola's flagship communications control processor. By RapidIO standards, PowerQuicc III is a big engine" makes it sound like Motorola bent over backwards for this 160,000 unit market. Yet they wouldn't take a financial risk in developing a G5 in an acceptable time frame for Apple and the potential for millions of sales.
I know I'm leaving a lot out concerning the MPC 8540/MPC8560 and PowerQuicc markets, but to me this is how the article reads.
Originally posted by Nevyn
See, this is where I get lost.
The G4s were the 7400, 7450, 7455, and now/soon the 7457 and 7457-RM.
The "cancelled" chip was the 7_5_00. Yes? The "G5"? The one rumored to be floating around in the "welded" boxes and "exploding"?
The (again, rumored) 7500 was apparently cancelled as well. I was responding to a poster above who seemed to say that the 7457 would be cancelled as well. As THT has mentioned, it's not been cancelled, just postponed for too long. The rumors were that the memory-controller-sporting, RapidIO-enabled version of the '57 was a goner. As well as the 7500.
I could be confused, but the 7457 doesn't seem like a 'generational change' sort of chip. More like the way the G3 has progressed to smaller/cooler/whatnot. It's still a G3.
If the 7457 is a 7455 with a die shrink and a bigger cache, and the 7457-RM is a 7457 with a few performance enhancements and an on-die memory controller with a RapidIO interface, then we're not talking about generational changes. The 7457 and the 7457-RM are just G4s. The 7500 was something else altogether, if the few scraps of rumors about it were right.
Mot won't make what they can't sell, but (again as THT points out) they also can't sell what they can't make. That's probably where everything went wrong.
Originally posted by rickag
Another thing that struck me funny is the reference to the potential "160,000-unit-per-year" "cellular-basestation market". This quote, "Motorola's flagship communications control processor. By RapidIO standards, PowerQuicc III is a big engine" makes it sound like Motorola bent over backwards for this 160,000 unit market. Yet they wouldn't take a financial risk in developing a G5 in an acceptable time frame for Apple and the potential for millions of sales.
That 160,000 unit market is just one target for the PowerQuicc III -- there many many many others, if they can deliver their product. The earlier PowerQuicc chips were a multi-billion dollar revenue generator for Motorola, and they're hoping the new ones will be too (if they ever get them out the door). They are probably right too because they are great chips for the markets they are aimed at.
As Amorph said, one of the key features of the PowerQuicc III design is rapid customization / modularity. Moto didn't have to bend over backwards to address a particular need in this market, they just had to lean a little to the side and none of their other customers would even notice. Developing a G5 (or G4+RIO) is a much more expensive undertaking with only one customer (at least for the G5, they might have others for a G4+RIO which is why they still mention it occasionally to see if anyone asks for it).
Originally posted by Programmer
That 160,000 unit market is just one target for the PowerQuicc III -- there many many many others,.......
Programmer, you're absolutely correct.
That's why I said,"...makes it sound like Motorola bent over backwards for this 160,000 unit market."
I also said,"I know I'm leaving a lot out concerning the MPC 8540/MPC8560 and PowerQuicc markets, but to me this is how the article reads." I really did understand the market is much larger than 160,000.
It just struck me funny, I guess my sense of humor is somewhat warped. For years now, I've only had one gripe w/ Motorola. My gripe pertains to their seeming unwillingness in the past 3 years to take a chance on Apple and develop a cpu specifically designed for a desktop, not just make the G4 OK for a desktop The G4 has fallen, arguably, woefully behind the competition.
Oh the irony, of years of hoping Motorola would take that chance, then reading an article making it sound like Motorola bent over backwards for 160,000 made me laugh(out loud at that).
It would have behooved the author of the article to mention the total potential market for these processors not just the cellular base station market,...
And it would have behooved me to be clearer in my post.
I've never seen the contract between Apple and MOT so I don't know what requirements it incorporated. Apple may have pulled a non-performance clause out of the contract and said "give us this processor performance by a certain date or we're looking elsewhere".
Originally posted by Bigc
Was MOT's problem that they didn't want to "take a chance" with Apple or was it that they couldn't get their Fabs to work to guarantee delivery to Apple. Recent info seems to suggest the latter.
Motorola have many more problems than an inability to get their new process working. They are desperately short of chip designing talent, and have real problems keeping engineers at the company. The company is notorious as a bad place to work, Dilbert's early columns were based on employee experiences at the company!
The SPS is perenially underfunded (thus the problems with HiP 7), you can't do semiconductors on the cheap. When you cut budgets your best talent often walks, into better jobs at Intel, IBM and, most frequently, AMD. If you don't have senior chip engineers around you also loose experience, your design ethos and methedologies and historical perspectives. Worse, you loose the ability to train new engineers in the practical methods of design.
Senior engineers are at the "tiller" of the design process. They make sure the general direction of the project is kept on course, deadlines are met and problems are fixed. Without that level of design experience it is extremely difficult to deliver a chip on-time, on-budget and within the target design parameters (frequency, performance, die area and power draw being the major ones).
The 7500 project was outside of all parameters, late into the design cycle with major attrition of the engineers working on it.
Motorola made the only desicion they really could. They were in no position to bet the company on this late and expensive behemoth, so it got dropped and the remaining engineering talent redistributed to other achievable projects.
For the record:
I can confirm that the 7457-RM is no longer on the G4 roadmaps from Mot.
[Edit: Oh, and I can also confirm that it WAS on the roadmaps before. But you may have known that already...]