cait-sith said:
No heat problem, it was the sensor.
[I feel like John Belushi. Instead of "no Coke, Pepsi", it's "no heat, sensor".]
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
PowerMac G5 Cooling/Sensor Issues?
- Thread starter MacRumors
- Start date
- Sort by reaction score
No heat problem, it was the sensor.
[I feel like John Belushi. Instead of "no Coke, Pepsi", it's "no heat, sensor".]
blackfox
macrumors 65816
Not that I don't think your point about the tier structure has validity, but just what would you suggest in place of it? Are you complaining about the use of the structure itself or Apples' implementation of it? Even w/o speed bumps to the PM (b/c of chip issues), would an implementation of G5s in the upper-ranged iMacs and PBs satisfy?(say 1.4 - 1.6 ghz)Personally I think Apple could buy itself some time here...assuming they can figure out cooling issues for spaces that small...anyway, just curious...thanksDont Hurt Me said:
Dont Hurt Me
macrumors 603
I agree, it would have cost less money and they would allways be running with the PC crowd and wouldnt have to worry about "special video cards" or failed Cpu makers. Plus they wouldnt have to get companies to make a special Mac version of software. Following the last 5 years of motorola has shown this was a mistake. I still think this was one of the worst moves ever made sticking with Moto & LowPowerPC. Im sure IBM will work it out but a Year of stagnation after all those motostagnation years is just very very sad for the Mac user.Frisco said:
does anyone know a good dentist? I think steve needs one since he's kicking himself for putting his foot in his mouth last year.
maybe they'll add more processors instead of faster ones? a nice quad-g5 system would make everyone happy, even if they can't hit 3ghz.
maybe they'll add more processors instead of faster ones? a nice quad-g5 system would make everyone happy, even if they can't hit 3ghz.
Processor issues.
The speed issues that IBM are having make more sense if you give credit where it's due. Chips are hard to make and Intel make more of them than anyone (microprocessors - not memory chips). They're really - really good at it. They still run 20 year old code with all its design limitations as fast as the best clean-sheet designs their competitors are capable of.
I'm a car fan. Though there's something beautiful about small manufacturers like Ferrari and Apple they have real problems until they hook up with the biggest suppliers and focus on what they do best (which in Apples case is industrial design and interfaces.)
I'd love to see Apple use Intel in a parrallel line of OSX machines. With all things being equal - and the ability to swap out a chip whenever Intel or AMD released a speedbump - I know a lot of people would go for them.
The superficial superiority of RISC architecture is a niche desire. For the PhotoShop and Word crowd (that's me) - it makes little real difference.
OR - revamp the architure to handle up to 8 or more processors. We could build little grids and get power that way. Music software's doing this now - graphics software's allowed it for ages. As a company that's aimed at creative Pro's it seems obvious. I'd buy an 8-way G4 with two factory processors for $1800 in a heartbeat. Charge me a couple of hundred for each new brain and I'd hand over another $1,200 with a smile.
Finally - IBM's problems are probably something dumb like air quality, power supplies or the qualiy of their silicon crystals. They'll fix it. Chips are tiny. If their yeilding any the technology's sound - it's the process they need to work on.
form said:
The speed issues that IBM are having make more sense if you give credit where it's due. Chips are hard to make and Intel make more of them than anyone (microprocessors - not memory chips). They're really - really good at it. They still run 20 year old code with all its design limitations as fast as the best clean-sheet designs their competitors are capable of.
I'm a car fan. Though there's something beautiful about small manufacturers like Ferrari and Apple they have real problems until they hook up with the biggest suppliers and focus on what they do best (which in Apples case is industrial design and interfaces.)
I'd love to see Apple use Intel in a parrallel line of OSX machines. With all things being equal - and the ability to swap out a chip whenever Intel or AMD released a speedbump - I know a lot of people would go for them.
The superficial superiority of RISC architecture is a niche desire. For the PhotoShop and Word crowd (that's me) - it makes little real difference.
OR - revamp the architure to handle up to 8 or more processors. We could build little grids and get power that way. Music software's doing this now - graphics software's allowed it for ages. As a company that's aimed at creative Pro's it seems obvious. I'd buy an 8-way G4 with two factory processors for $1800 in a heartbeat. Charge me a couple of hundred for each new brain and I'd hand over another $1,200 with a smile.
Finally - IBM's problems are probably something dumb like air quality, power supplies or the qualiy of their silicon crystals. They'll fix it. Chips are tiny. If their yeilding any the technology's sound - it's the process they need to work on.
Dont Hurt Me
macrumors 603
crippling game is no way to sell computers, if you have to have seperation do it with qty of CPUs and features slots etc. cant tell you how many people wouldnt even look at Imac or Emac because there isnt a videocard,pci slot or forces monitor on them so Apple tells these folks you Must buy a Powermac and then they do this thing again by not allowing a person to get the fastest cpu. If they do they must get 2 and a bigger harddrive so they Force them up the tier or, or they go buy a PC. marketshare tells us this is what is going on. Apple in essence pushed them away. Have a pro line and consumer line and let the user configure the machine. stop the crippling and stop the forcing and bend to the consumers or pro's needs. its really simple.blackfox said:
iMeowbot
macrumors G3
bennetsaysargh said:
Placement is everything for those parts. If the CPU assembly is a different shape or mounted dofferntly than in the older model, the warmer and cooler spots will be different, and it can take a lot of testing to figure out what the best points are (& how many points need to be measured) to obtain useful readings. Then, the software needs to be modified to do something useful with the data (what fans get turned on, what speeds, blah blah).
rdowns
macrumors Penryn
blackfox said:
I'll bite. It is common belief that Apple won't upgrade their iMac and eMac lines so they don't come too close in performance to the PM line. Personally, I coin this "head up your ass" marketing. There is already enought diffrentiation in the Imac vs. PM line to put G5s in iMacs....
Dual vs. single processor
RAM slots
Expansion slots
FireWire 800
Drive bays
Video card
Could do slower optical drives in iMac- 4x vs. 8x
Hard drive size
In my mind, that is plenty to diffrentiate the lines. If you need expandability/upgradability, go PM, if not, go iMac.
As for the notion that a G5 can't go in the iMac before the PB, also "head up your ass" marketing. The users and markets for both are dissimilar. Desktop vs. portable.
Does this impact new Powerbooks?
What are your thoughts on this speed/heat issue as far as the PowerBook updates...which is what many of us are truly waiting for?
What are your thoughts on this speed/heat issue as far as the PowerBook updates...which is what many of us are truly waiting for?
Dont Hurt Me said:
Well, we would still need a special version of the software (OS X not Windows).
I am glad Apple didn't go with Intel at the time since we would of ended up with the Itanium2 which according to Intel's website runs at 1.50 GHz, 1.40 GHz, 1.30 GHz (http://www.intel.com/products/server/processors/server/itanium2/index.htm) and from recent news on Intel's change of heart about 64-bits might have left us stranded.
AMD would have been a little more promising with 2.2 GHz (http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_8796_9240,00.html) although we would need the low power 30w version (EE).
Solaris and PM's future
I've always liked Solaris. It's a pity Sun hasen't being doing that well lately but that 1.6 Billion $US from Microsoft should help.
I've always thought that just as Apple and Sony would make a good team on the entertainment gadget front, Apple and Sun would make a good team on the Unix enterprise front. If only Adobe would port FrameMaker to OSX!!
With respect to the future of the PM, I'm wondering if we've simply hit a technology wall at dual 2.0. If so, eventually Apple and their suppliers will overcome it.
In the meantime perhaps the solution lies in distributed processing. As someone mentioned in another thread, they were in no hurry to purchase a new PM tower. They would rather build a cluster of headless XServes.
What I'm suggesting is that instead of sinking all of their R&D into a faster standalone box, Apple may serve us better by making existing technology cheaper. Imagine an XServe cluster where adding an additional processor was as cheap as adding memory is now. That individual processor wouldn't have to be fastest available, just cheap so even the home user could afford them in quantity.
Just thinking out loud, please don't beat me up or be rude. 🙂
~iGuy
cait-sith said:
I've always liked Solaris. It's a pity Sun hasen't being doing that well lately but that 1.6 Billion $US from Microsoft should help.
I've always thought that just as Apple and Sony would make a good team on the entertainment gadget front, Apple and Sun would make a good team on the Unix enterprise front. If only Adobe would port FrameMaker to OSX!!
With respect to the future of the PM, I'm wondering if we've simply hit a technology wall at dual 2.0. If so, eventually Apple and their suppliers will overcome it.
In the meantime perhaps the solution lies in distributed processing. As someone mentioned in another thread, they were in no hurry to purchase a new PM tower. They would rather build a cluster of headless XServes.
What I'm suggesting is that instead of sinking all of their R&D into a faster standalone box, Apple may serve us better by making existing technology cheaper. Imagine an XServe cluster where adding an additional processor was as cheap as adding memory is now. That individual processor wouldn't have to be fastest available, just cheap so even the home user could afford them in quantity.
Just thinking out loud, please don't beat me up or be rude. 🙂
~iGuy
Frisco said:
I totally agree. When you're the only major buyer of a chip, the chip manufacturer can put you on the back burner if they feel it's not worth their time (Motorola). The advantage of Intel and AMD is: That is their business! If they don't produce newer and faster chips, they go out of business! There is competition, and there is a will to sustain. Isn't that how economy works?
To the person commenting on how Apple DOESN'T have trouble competing with their OS:
A major factor why people don't buy Macs (besides hardware), is the lack of software or incompatibility of software they've already bought. This is my also my biggest concern. I know Apple has the basic stuff, but when you search google for a program to do something, you can usually find it for Windows. The software for d/ling programs to my phone doesn't exist for macs. I think this is a problem. Perhaps if apple was more competitive, and had a bigger marketshare, more people would write programs for Macs.
How many people here could live without touching a PC? That is, no gaming rigs, no specialty software (game cracks 🙂 ), and very little freeware. Until 90% of you can say yes, I think Apple has some work to do.
avkills
macrumors 65816
Dont Hurt Me said:
You have no idea how computers work do you?
For one, yes the video cards would have to be different because Apple would not be using stock motherboards and they also like designing their own system controller chips. ATI and nVidia could make video cards that work in both, but they don't because it does not make sense money wise. Sure they make Mac cards, but the bleeding edge is with PCs, so that is what they focus on first. Case in point, Black Magic Design makes PCI based uncompressed SD cards that work in both Mac and Windows.
Software would have to be written specifically for OS X's API sets. You think magically going to x86 or AMD and we could install windows software, uhhh no. Windows and OS X use completely different API sets.
Carry on.
-mark
The yield issue is confirmed by the fact that 2 months ago Digitimes reported that Nvida was moving production from IBM back to TSMC because of yield issues.
Penman said:
Just because you produce a lot of a product doesn't mean you are any good at it. Look at GM and Ford. Just beacause Intel makes a lot of chips doesn't mean they produce the fastest/best chips. Intel's best new chip, the Pentium M wasn't even designed in-house.
BTW, RISC design is not a "superficical superiority". All modern X86 CPUs have a huge instruction decode process to turn the original crappy mem-to-mem operands of the X86 asm into a pseudo RISC instruction stream. Without it, they'd never be able to sustain the pipeline depth that they currently enjoy. RISC designs allowed for a far better use of multiple parallel execution units, which is how the current CPUs get superscalar execution that is not normally possible if you're using the original X86 instructions.
The problem is software: Most software developers nowadays are barely capable of understanding true parallel execution theory, much less the nuance of implementing it correctly. This is why its only used in very specialized cases.
For the typical computer "user", most issues of "percieved speed" are absolutely gone given today's hardware with enough memory and gigahertz class CPUs. If the UI is responsive, the computer is "fast", which is an area where Apple currently blows the pants off its nearest OS rivals.
I agree that IBM will fix yields and speed in the near future. They *were* the first to implement this process on this scale, so they were bound to have some kinds of issues to iron out. It'll get better.
rdowns said:
It didn't stop them from releasing Itanium 2 and the wonderful i960 back in the day... 😀
blah blah blah..... heat this, aluminum shortage that....all i hear is excuses. Just make the damm thing like dual 5 Ghz,, a wind tunnel, and armor plate with steel. Ship some Bose Noise Canceling headphones with it and make it an awesome promotion and just give us what we want. Let the people around you using windows machines go deaf. Who cares, atleast we will be happy with our pmacs. Stop making stupid excuses apple.
Frisco said:
First off, if you change CPUs, yet again, you break binary compatibility and throw away all the time your software manufacturers spend optimizing to your platform. OSX is nice because it works only on Apple hardware, which is touted as having far fewer defects/problems than most mass-produced crap from Dhell and others. Plus, they'd be under serious pressure from the consumer market to release OSX for hardware which is not theirs and they'd subsequently go out of business. Tell me how well Solaris sells for X86 hardware: I thought so. As the story goes: Good is the enemy of Excellent.
They are a niche company, selling SOLUTIONS to their customers. That's why they've gotten into the application market. Make the best product possible, that runs on very specific hardware that you sell. If its better than the competition, and you get buy-in, then you have a sustainable business model that works in the long-run. If you're just hardware or just software, you lose. IBM figured this out early-on with the mainframe with CICS and IMS running on MVS. A complete end-to-end business solution for a specific markets that has no real competition, even today...
Different hardware is GOOD for Apple, trust me...
SiliconAddict
macrumors 603
Ummm this sucks. But once again Apple is held by the throat of one company. In this case IBM. As much as I and everyone else were hoping that IBM would bring about a fast and furious hardware campaign it looks like, at least for now, that dream is a dud.
I'm wondering what is the possibility of AMD designing a PowerPC because unless IBM gets their collective act together soon they are going start looking like Moto. I'm not going to scream Moto-itus just yet but its not looking good. 🙁
I'm wondering what is the possibility of AMD designing a PowerPC because unless IBM gets their collective act together soon they are going start looking like Moto. I'm not going to scream Moto-itus just yet but its not looking good. 🙁