The only thing I can see them doing, and this is IF they can get a hold of the companion chip AND 970 itself, is to make a single 970 upgrade card for Sawtooth type G4 machines that includes it's own integrated DIMM slots. DIMM slots can be mounted off to the side and you would have to remove ALL RAM from the motherboard unless the card can somehow use it as disk cache or something. That would mean they would have to make a controller that connects the companion chip on the daughtercard to the main controller on the motherboard. Basically a HyperTransport to MPX bridge. Its possible but would require some effort on the upgrade manufacturers part. Any solution that uses the RAM on the mobo would negate most of the G5 advantages.
This would be an expensive upgrade too. Maybe that is why PL is so pissed. Oh well.
The only thing I can see them doing, and this is IF they can get a hold of the companion chip AND 970 itself, is to make a single 970 upgrade card for Sawtooth type G4 machines that includes it's own integrated DIMM slots. DIMM slots can be mounted off to the side and you would have to remove ALL RAM from the motherboard unless the card can somehow use it as disk cache or something. That would mean they would have to make a controller that connects the companion chip on the daughtercard to the main controller on the motherboard. Basically a HyperTransport to MPX bridge. Its possible but would require some effort on the upgrade manufacturers part. Any solution that uses the RAM on the mobo would negate most of the G5 advantages.
This would be an expensive upgrade too. Maybe that is why PL is so pissed. Oh well.
You are right, it's the only way they could do this stuff. Anyway it will be expansive and it will perform badly for all the I/O stuff. I think it will be better to sell an old mac, and buy a new one, than buy an upgrade.
One thing for sure, I'm going with a Gigadesign G4 upgrade later this year to my poor old 400MHz G4. I figure a 1.x GHz G4 upgrade can tide me over for a few years until I have the dough for a REAL POWERmac. By then I'll probably be able to buy a dual 3.5 GHz monster....I can't wait!
The golden age stopped with the introduction of the G4 sawtooth and the Mpx bus.
Actually due to Motorola's lack of progress with the G4, it can make sense to upgrade a Sawtooth-era Mac. The motherboard on a G4 from 2003 isn't much better than one from 2000; just a 167MHz bus instead of 100. I got a Powerlogix 1.2GHz processor upgrade and a Radeon 8500 for my G4/400, and performance is as good as a recently-current G4.
The G5 will definitely be a problem for upgrade manufacturers. I wouldn't be suprised if somebody figures out how to replace a G5 processor with a faster one, but I doubt upgrades for G4s will be feasible.
Man, you can tell Apple has hit on something big with the G5. The amout of negative press its getting is the surest sign yet. Particularly as nobody has event tested one yet!
This is only one benchmark, of course...so one shouldn't draw comparisons until the G5s actually are shipping. Hopefully real G5 performance data will be available soon.
All the that PL says is that if the Skidmarks GT benchmark is an indication than the G5 is not worth the price difference. Nothing wrong with that, or?
In all my past jobs dealing with PC and Mac support, I have yet to perform a PC CPU upgrade or a Mac CPU upgrade for a production machine except for a few 7500 to G3 upgrades. Big business will just shell out for new machines if the old ones are getting to slow. The only worthwhile upgrade is RAM. Then upgrading a CD player to a CDRW or DVD-R if needed. If a HD goes then we replace that. CPU is the last and by then it makes more sense to upgrade the old machine. Point is, the G5 processor not being upgradable won't be the limiting factor in being adopted by business. PL served the home consumer (like me) that needs to pinch every penny.
Actually due to Motorola's lack of progress with the G4, it can make sense to upgrade a Sawtooth-era Mac. The motherboard on a G4 from 2003 isn't much better than one from 2000; just a 167MHz bus instead of 100. I got a Powerlogix 1.2GHz processor upgrade and a Radeon 8500 for my G4/400, and performance is as good as a recently-current G4.
The G5 will definitely be a problem for upgrade manufacturers. I wouldn't be suprised if somebody figures out how to replace a G5 processor with a faster one, but I doubt upgrades for G4s will be feasible.
You are right i should have been more precise. Upgrading a sawtooth design worth the price. But upgrading a beige or a B&W G3 with a G4 is quite disapointing.
What i meant that when they have introduced a new mobo aka the sawtooth, the upgrade of previous generation of mac where worthless.
The only thing I can see them doing, and this is IF they can get a hold of the companion chip AND 970 itself, is to make a single 970 upgrade card for Sawtooth type G4 machines that includes it's own integrated DIMM slots.
This would be an expensive upgrade too. Maybe that is why PL is so pissed. Oh well.
This commentary is an outrageous and transparent attempt to save your threatened upgrade business. I won't begin to get into a point by point rebuttal of your commentary, though one has to look no further than the first paragraph of the "article" to see the wishful thinking being displayed ("...whether to keep their existing G4 and upgrade with a CPU card from PowerLogix; or buy a new G5 and hand down their previous PowerMac G4 to someone else in the office...perhaps installing a CPU upgrade in it at the same time."). Please!
This is about one thing and one thing only: Should the G5 take hold as the Mac Pro machine of choice (as it will), your company will have nothing to sell. It's that simple. The article itself is utterly tasteless. You are bashing an unreleased machine with the hope that it benefits your business. You are also flat-out lying. I bring to your attention this quote:
"However 82% of the performance for about 28% of the actual out of pocket cost is a bargain by anyone's measure. This is only one benchmark, of course...so one shouldn't draw comparisons until the G5s actually are shipping. "
Your 1.4g Dual card costs $999. Apple's top-of-the-line G5 will cost $2999. That's 33%, not 28%. With the G5, we get an entirely new system architecture, much faster bus speeds, etc. We can also get a low end G5 for just $1999. Granted, it's more expensive. As far as not drawing comparisons, that is EXACTLY what you have attempted to do...using one benchmark. Even WITH this benchmark, your machine gets walloped by almost 30%. Then, you coyly add: "[Of course, we're not really making any judgments yet...*cough*]".
Coming from a company with an obvious agenda, the commentary is totally invalid. I've lost a great deal of respect for your company based on this one article alone. I'd think it would be better to focus your energies on positive marketing for your existing line, and perhaps if feasible, finding a way to get into the G5 upgrade business in time.
I appreciate your opinion, but there will indeed be many things for us to sell...we already have 970 designs in progress.
The point we made is quite simple: wait til the machines are out before making a decision as to how fast they are. That's all. We want apple to sell G5s...don't forget that's one of the earliest comments in the article.
So, I sent him this:
"Thank you for responding. I am glad you have 970 designs in progress.
What I am saying (and apparently what a lot of other users are saying) is that your company's criticisms of an unreleased product are perceived to be inappropriate and self-serving. As far as wanting Apple to sell G5's and the first paragraph of the article, I have to question the company's motives. Why would the company want this? There isn't going to be a real G5 upgrade market (unless you can manage a G4 to G5 upgrade) for perhaps 18-24 months. G5 machines will not be good for PowerLogix. There is no question about that. As I said, the article is transparent and conveys wishful thinking. I am sorry, but somehow I doubt the majority of users will opt to drop $999 on a 4 year old (for example) G4 system upgrade when they can get a new 64 bit system for $1999. PowerLogix knows this and sanctioned this article which feigns impartiality while exposing its rather obvious motives.
And finally, going after rabid Color Classic upgraders, the G5 Color Classic kit includes a whole new slide-in motherboard and back cover, along with some new monitor circuitry to turn the 10" display into a multisync. Onboard video will support both the internal monitor (at 512 x 384, 640 x 480, or a somewhat fuzzy 800 x 600) as well as an external display.
Some people are probably already plotting the Color Classic G5. There are Color Classic G4s (using Cube motherboards), but they'll probably have to wait for G5 iMacs/eMacs for the CC G5.
I'm so getting one. Sidenote: I wonder how much a barebones Cube on the lowest end goes for. I have a Color Classic and I am just crazy enough to do that. For real. It's a very small case with a nice CRT. A hack enables 800x600 on it. However maybe I'll just scavenge for a small LCD, with 1024 by 768 capability. Then I'll finally have a nice desktop (get a 2nd monitor for it of course) and nice laptop. Giggity giggity!
I felt the same urge to write them a "pish off" letter, but it's not worth the caloric expenditure required SDW. They are basically trying to keep people on the low end from buying 1.6 and 1.8 GHz G5's, suggesting instead that they upgrade their current G4 to 1+ GHz with their cards. That is the only way they will stay financially healthy for the next 12+ months.
That whole article is so transparent it's ridiculous. Almost laughable. You can see them thinking their hatchet-job through, trying to make it sound like "unbiased product literature". The whole thing makes me glad I bought a Sonnet card and not a PowerLogix card. Hopefully the former will have enough good taste to just keep their mouths shut and continue finding other revenue streams for the time being.
Obviously, there most likely will never be a G5 upgrade for G4 and earlier era machines. Only for G5 machines. Hence, nobody will be looking to buy one for at least 18 months or so. And it will take them that long or more just to design and develop a solution for those people.
I saw some (legit) Cube logic boards on eBay a while ago for £50, which is becoming more of a bargain as RAM/G4 ZIF card/reasonable video card/etc.. prices fall. Might be worth watching out for. A dual head video card (Radeon 7000?) might be ideal. The trickiest thing is probably getting a decent PSU.
Comments
This would be an expensive upgrade too. Maybe that is why PL is so pissed. Oh well.
Originally posted by Outsider
The only thing I can see them doing, and this is IF they can get a hold of the companion chip AND 970 itself, is to make a single 970 upgrade card for Sawtooth type G4 machines that includes it's own integrated DIMM slots. DIMM slots can be mounted off to the side and you would have to remove ALL RAM from the motherboard unless the card can somehow use it as disk cache or something. That would mean they would have to make a controller that connects the companion chip on the daughtercard to the main controller on the motherboard. Basically a HyperTransport to MPX bridge. Its possible but would require some effort on the upgrade manufacturers part. Any solution that uses the RAM on the mobo would negate most of the G5 advantages.
This would be an expensive upgrade too. Maybe that is why PL is so pissed. Oh well.
You are right, it's the only way they could do this stuff. Anyway it will be expansive and it will perform badly for all the I/O stuff. I think it will be better to sell an old mac, and buy a new one, than buy an upgrade.
Originally posted by Powerdoc
The golden age stopped with the introduction of the G4 sawtooth and the Mpx bus.
Actually due to Motorola's lack of progress with the G4, it can make sense to upgrade a Sawtooth-era Mac. The motherboard on a G4 from 2003 isn't much better than one from 2000; just a 167MHz bus instead of 100. I got a Powerlogix 1.2GHz processor upgrade and a Radeon 8500 for my G4/400, and performance is as good as a recently-current G4.
The G5 will definitely be a problem for upgrade manufacturers. I wouldn't be suprised if somebody figures out how to replace a G5 processor with a faster one, but I doubt upgrades for G4s will be feasible.
1.2 ghz G4 upgrades are now going for $450 and dropping FAST.
Daddy's G4@450 tower is going to be getting a new processor soon. OWC's got some great prices and deals.
Nick
This is only one benchmark, of course...so one shouldn't draw comparisons until the G5s actually are shipping. Hopefully real G5 performance data will be available soon.
All the that PL says is that if the Skidmarks GT benchmark is an indication than the G5 is not worth the price difference. Nothing wrong with that, or?
End of Line
Originally posted by 3.1416
Actually due to Motorola's lack of progress with the G4, it can make sense to upgrade a Sawtooth-era Mac. The motherboard on a G4 from 2003 isn't much better than one from 2000; just a 167MHz bus instead of 100. I got a Powerlogix 1.2GHz processor upgrade and a Radeon 8500 for my G4/400, and performance is as good as a recently-current G4.
The G5 will definitely be a problem for upgrade manufacturers. I wouldn't be suprised if somebody figures out how to replace a G5 processor with a faster one, but I doubt upgrades for G4s will be feasible.
You are right i should have been more precise. Upgrading a sawtooth design worth the price. But upgrading a beige or a B&W G3 with a G4 is quite disapointing.
What i meant that when they have introduced a new mobo aka the sawtooth, the upgrade of previous generation of mac where worthless.
Originally posted by Outsider
The only thing I can see them doing, and this is IF they can get a hold of the companion chip AND 970 itself, is to make a single 970 upgrade card for Sawtooth type G4 machines that includes it's own integrated DIMM slots.
This would be an expensive upgrade too. Maybe that is why PL is so pissed. Oh well.
hmmm... deja vu.
dropping a G5 mobo upgrade into a IIci
although i could totally see the NeXTCube free swap. Magnesium better than Al for cooling?
Dear Sir or Madam,
I must take issue with your comments on the G5 introduction. (http://www.powerlogix.com/press/rele...03/030625.html).
This commentary is an outrageous and transparent attempt to save your threatened upgrade business. I won't begin to get into a point by point rebuttal of your commentary, though one has to look no further than the first paragraph of the "article" to see the wishful thinking being displayed ("...whether to keep their existing G4 and upgrade with a CPU card from PowerLogix; or buy a new G5 and hand down their previous PowerMac G4 to someone else in the office...perhaps installing a CPU upgrade in it at the same time."). Please!
This is about one thing and one thing only: Should the G5 take hold as the Mac Pro machine of choice (as it will), your company will have nothing to sell. It's that simple. The article itself is utterly tasteless. You are bashing an unreleased machine with the hope that it benefits your business. You are also flat-out lying. I bring to your attention this quote:
"However 82% of the performance for about 28% of the actual out of pocket cost is a bargain by anyone's measure. This is only one benchmark, of course...so one shouldn't draw comparisons until the G5s actually are shipping. "
Your 1.4g Dual card costs $999. Apple's top-of-the-line G5 will cost $2999. That's 33%, not 28%. With the G5, we get an entirely new system architecture, much faster bus speeds, etc. We can also get a low end G5 for just $1999. Granted, it's more expensive. As far as not drawing comparisons, that is EXACTLY what you have attempted to do...using one benchmark. Even WITH this benchmark, your machine gets walloped by almost 30%. Then, you coyly add: "[Of course, we're not really making any judgments yet...*cough*]".
Coming from a company with an obvious agenda, the commentary is totally invalid. I've lost a great deal of respect for your company based on this one article alone. I'd think it would be better to focus your energies on positive marketing for your existing line, and perhaps if feasible, finding a way to get into the G5 upgrade business in time.
Thank you.
I appreciate your opinion, but there will indeed be many things for us to sell...we already have 970 designs in progress.
The point we made is quite simple: wait til the machines are out before making a decision as to how fast they are. That's all. We want apple to sell G5s...don't forget that's one of the earliest comments in the article.
So, I sent him this:
"Thank you for responding. I am glad you have 970 designs in progress.
What I am saying (and apparently what a lot of other users are saying) is that your company's criticisms of an unreleased product are perceived to be inappropriate and self-serving. As far as wanting Apple to sell G5's and the first paragraph of the article, I have to question the company's motives. Why would the company want this? There isn't going to be a real G5 upgrade market (unless you can manage a G4 to G5 upgrade) for perhaps 18-24 months. G5 machines will not be good for PowerLogix. There is no question about that. As I said, the article is transparent and conveys wishful thinking. I am sorry, but somehow I doubt the majority of users will opt to drop $999 on a 4 year old (for example) G4 system upgrade when they can get a new 64 bit system for $1999. PowerLogix knows this and sanctioned this article which feigns impartiality while exposing its rather obvious motives.
Good Luck.
Originally posted by curiousuburb
hmmm... deja vu.
dropping a G5 mobo upgrade into a IIci
although i could totally see the NeXTCube free swap. Magnesium better than Al for cooling?
That article is pretty darned neat... and completely insane.
Ass.
And finally, going after rabid Color Classic upgraders, the G5 Color Classic kit includes a whole new slide-in motherboard and back cover, along with some new monitor circuitry to turn the 10" display into a multisync. Onboard video will support both the internal monitor (at 512 x 384, 640 x 480, or a somewhat fuzzy 800 x 600) as well as an external display.
Some people are probably already plotting the Color Classic G5. There are Color Classic G4s (using Cube motherboards), but they'll probably have to wait for G5 iMacs/eMacs for the CC G5.
That whole article is so transparent it's ridiculous. Almost laughable. You can see them thinking their hatchet-job through, trying to make it sound like "unbiased product literature". The whole thing makes me glad I bought a Sonnet card and not a PowerLogix card. Hopefully the former will have enough good taste to just keep their mouths shut and continue finding other revenue streams for the time being.
Obviously, there most likely will never be a G5 upgrade for G4 and earlier era machines. Only for G5 machines. Hence, nobody will be looking to buy one for at least 18 months or so. And it will take them that long or more just to design and develop a solution for those people.
Originally posted by SDW2001
The guy responded again. He's now getting into point by point rebuttal.
Ass.
He must read AppleOutsider.