Who here has overclocked macs?

[Mord]

No need to be so harsh.

Unless anyone here is an insider for the semiconductor devision of IBM we cannot be sure. It isn't that uncommon of a practice for chip makers to "clock" chips faster (among other minor changes) to get faster speeds from them.

It is the act of things like adding more cache or a minor instruction change that will let them "get away" with calling them new chips.

But then again I am on the outside of this. A guy that may know would be Aidenshaw, but he never checks these threads .

the very definition of overclocking is someone other than the manufacturer setting the clock speed higher, these chips are run at their rated by IBM speed, i just hate the way people make assumptions like if it's watercooled it's overclocked, or if motorola does no list it it must be overclocked and then spread it around like they know better and thier word is fact.

 

[840quadra]

the very definition of overclocking is someone other than the manufacturer setting the clock speed higher, these chips are run at their rated by IBM speed, i just hate the way people make assumptions like if it's watercooled it's overclocked, or if motorola does no list it it must be overclocked and then spread it around like they know better and thier word is fact.

No worries, and I understand your gripe, and the true definition of the term. I still believe that Manufacturers themselves speed up clock cycles to have a quick and easy way to have a "newer and faster" processor model. I am not sure if IBM does this, but mostly all other chip makers have done this in their history.

It may not be called "overclocking" when the manufacturer does it, but they still do cheat. Like I said, I am not sure nor can I prove that IBM does this, but I would guess (not post fact) that some of their recent offerings may be clock speed bumps of older models. Do I think the 2.7 is an overclocked 2.3? No, but I wouldn't be surprised if a 2.7 is a slightly modified and overclocked 2.5, but that's just me I guess.

 

[Mord]

let me explain how processors are manufactured.

IBM makes a disk of cpu's and predicts the quality of the processors, they used to hand speed bin them but it is very uneconomic to do it that way so they have to hand speed bin a few and then use the data from the hand binning to predict the quality of the processors, the cpu's on the outer edge of the disk are the worst quality and the cpu's in the middle are the best due to the silicone crystalline structure, their are two limits to the clock speed of these prossessors, the first is the physical length of the longest consecutive string of transistors, and that is usually the number of transistors in the execution unit divided by the pipeline length, though that data is only known to IBM (unless i did some tedious maths and studies die diagrams which tbh i cant be arsed to so right now) you then work out the time it takes the a signal to cross that distance (say 0.1m divided by 3x10^8m/s) divide one by that number and you have the maximum attainable clock speed in Hz, for the 90nm G5 i suspect it's about 3.2GHz.

the other limit is the voltage, signals do not look like the digital signals you imagine they are more like sin waves, the constant is the voltage required to register logic level 1 or 0 and if that is the same as the maximum supply voltage then the signal will only register at the top of the wave and there for will be at logic level 1 for a very short time, when the supply voltage is allot higher than the voltage required the transistor will be triggered much faster as it at the steep part of the curve and the signal will be allot clearer, and can be clocked allot higher without becoming unstable, so IBM needs to set the highest voltage the the processor can take while remaining stable, as the cpu's around the edge are more susceptible to electron tunneling and electromigration and they shorten the life of the processor as they eat away at the cpu and make let signals jump track, and high voltages make it worse, so IBM rates the cpu lifespan then applies the highest voltage that will comply with that then rates the clock speed according to the voltage, this is all predicted every time their is a respin (change in design/technique/tollerences)

IBM cannot overclock by the definition of the word, they can be too lax in their clock speed ratings, but we dont see widespread reports of G5 cpu failure do we?

 

[840quadra]

let me explain how processors are manufactured.

Well you seem very passionate about this subject, and I do respect that, and your knowledge in regards to this subject. However you are starting to take it too personally.

Thanks for the write-up and the lesson. I understand loosly (from tech doccuments from IBM, Intel and AMD) how processors are built, How instruction sets work, pipes, floating points, and even grasp Moors law to a point. I don't know enough to build one, fix one, I guess I understand enough to be dangerous.

What I still want to know, is the layout of transistors, Silicon, Copper, and the full internal architecture unique on the G5 2.5 as compared to the G5 2.7? Are the clock cycles of these chips controlled internally, or by something else?

And before you label me a "clock speed junky" please note that I push the following point to my friends when they start bragging about their Intel HT processors, or AMD friends. I also know (by watching past debates between Jiggie2g (sp?) that you agree with the following.

In the microprocessor world, performance is defined by throughput and capacity -- not just clock speed. Processor frequency, cache size, memory bandwidth, and processor architecture all contribute to overall performance. At the 2004 International Solid-State Circuits Conference, a panel of processor architects from IBM, AMD, Intel, Fujitsu, Sun, and Stanford University generally agreed that chips will increasingly rely on parallelism, rather than clock rates, for achieving faster speeds.

 

[Mord]

woah, i'm not emotionally involved, i just call people out when i see they are saying things that are not so.

as for the 2.7 compared to the 2.5 they are the same electronically, but the 200MHz bump is the result of a respin where they tweak a few things to get the cpu's to tape out at a better quality then revaluate the wafer to calculate yields.

while i know that architecture is more important i'm talking from a one architecture design standpoint where more clockspeed is always better, if we were to have a discussion on architecture i'd probably take half an hour or so to do all the research for one post, architecture is all about trade offs, the balance between clock speed and pipeline length, and you can optimize the correct balance for your architecture, aka for netburst (P4) i'm sure the 30 odd stage pipeline is the best considering the rest of the cpu and with AMD's hammer i'm sure the shorter pipeline they use and the lower clock speed is the best trade off.

 

[VAmin]

I wonder if I can overclock my old G3 iMac 400. Do you know how? Is it possible?

I overclocked my iMac DV 400 to 500, but it is difficult. I would not attempt it unless you have good skills with a soldering iron.

Instructions: http://www.bekkoame.ne.jp/~t-imai/imacde1.html

It ends up running pretty hot, but it's never been unstable for me. I also have the video card overclocked about 40% wtih Aticcelerator II.

 

[840quadra]

woah, i'm not emotionally involved, i just call people out when i see they are saying things that are not so.

No need to "call me out" as I stated many times I wasn't 100% sure what I was talking about, and am more or less looking to take on some information. I just wanted you to know that I wasn't personally going after you, as quite honestly, I agree with many of your posts (past and current). I often post up debate to learn more information, not to sound like a pro or arrogant, just so you know . Makes me sound like a know it all if you don't really read into what I am saying, but trust me, I am not a know it all (well I try to be with modern Volkswagen / Audi but that is a separate forum and topic )

as for the 2.7 compared to the 2.5 they are the same electronically, but the 200MHz bump is the result of a respin where they tweak a few things to get the cpu's to tape out at a better quality then revaluate the wafer to calculate yields.

Thanks for the info, but were is the clock speed actually set? Is that inside the 970 or is that part of the ASIC and or system board?

while i know that architecture is more important i'm talking from a one architecture design standpoint where more clockspeed is always better

Not to be rude, but that is quite obvious. I was more or less being preemptive to clear the impression that I was all about clock speed. If I was, I would have to change my Screen title to 840quadraG6 ( yeah that sounds bad )

 

[LimeiBook86]

I overclocked my iMac DV 400 to 500, but it is difficult. I would not attempt it unless you have good skills with a soldering iron.

Instructions: http://www.bekkoame.ne.jp/~t-imai/imacde1.html

It ends up running pretty hot, but it's never been unstable for me. I also have the video card overclocked about 40% wtih Aticcelerator II.

Whoa awesome! I was thinking of doing that on my iMac, but it gets hot enough as it is. I also was going to do the iBook Overclock for the Clamshell models, it seems that that guide doesn't work on the Special Edition 466mhz (Lime/Graphite) iBooks so know that before opening it up...wish I know that before I opened mine.

 

[FocusAndEarnIt]

I overclocked my iMac DV 400 to 500, but it is difficult. I would not attempt it unless you have good skills with a soldering iron.

Instructions: http://www.bekkoame.ne.jp/~t-imai/imacde1.html

It ends up running pretty hot, but it's never been unstable for me. I also have the video card overclocked about 40% wtih Aticcelerator II.

Wow!! Thanks!