You may have heard by now that AMD is launching a campaign to fight the MHz Myth.  Part of this campaign includes naming their processors with new names that sound like MHz speeds, but aren’t.  With this tactic, the company hopes to defocus people’s attention on MHz.  For instance, an Athlon 1.4 GHz chip might be named the Athlon 1700.

My question is, should Apple do the same?  Rather than buying an 867 MHz G4, would you rather buy a PowerMac G4 1700?

I wanted to post my thoughts separately in order to not make the setup question sound biased.  So:

This is a bunch of crap from AMD.  Sure, MHz don’t matter, but the company shure as heck didn’t say that when it was keeping pace with Intel on a MHz basis.  Also, this renaming scheme is not fighting the MHz Myth, it is instead an effort to lie to its customers while PLAYING on the MHz Myth!  The company will only be furthering the urban myth that MHz matters, and this ploy smacks of desperation.  Also, note that AMD will actually go so far as to *hide* the MHz rating in the Bios.  Lame.

This is not the true path to reeducating the public, and I can’t imagine that Apple would take such a path too.  This is David Nelson’s idea, but I would love to see processors rated and named according to their Gigaflop rating.  David’s working on a piece to that effect, but I thought it bore mentioning now. 

I agree Mhz rating should be out the door.  Only gigaflops should be stated.

- gopher

I’m not sure about this one. I like the idea of defocusing on MHZ by offering a different name, but I have to agree with you, Bryan, that using arbitrary names is disinformation to the public and offers no real value.

Unless the new names are tied to something meaningful, say the gigaflop rating you propose, then Apple should not follow AMD in this scheme.

The whole problem with picking a new standard is that all the players have to agree on it. Intel will not likely want to use another rating system because that would mean they would have to shift gears and make processors that put emphasis on the new rating, as they do now with MHZ.I can see that they would come out looking pretty mundane compared to other processors if they did this.(Why be fair when you have the upper hand?)

Of course, Apple could force the issue. They could make sure that PC benchmarks run on Macs, hire an independent testing company to run the benchmarks against Macs and PCs, then publish the results with overall gigaflop ratings. Then publicly invite Intel or Dell to do the same. It might work. From there, any new processor could have a name that relates to its gigaflop rating. (G4-5.5 would be a G4 rated at 5.5 sustained gigaflops, for instance.)

It’s a system I’d like to see, but doubt we will.

When I talk of my Powerbook - I tell that it runs at 33,000 khz.

OOOPS I forgot a 3, that is 333,000 khz. Even better.

The word Gigaflop(s) doesn’t have the gutsy sound of MEGAhertz, but I agree with the complexity of today’s components a streamlined rating of system performance is needed.  When you buy a new kitchen appliance it has an energy rating. While this is helpful, refrigerators only do one thing preserve food using as little energy as possible.  Perhaps computers could use several categories such as: Spreadsheet, Graphics, and Gaming calculations.  I don’t think speed geeks will ever let go off the megahertz myth but the general public may welcome ratings using everyday understandable language.  This computer has a 5 star rating for business use, or this machine has only 3 stars for gaming but that’s fine since I hate playing games.  There is a big marketing strategy to be figured out here.

I don’t like the idea of “names” for processors. It’s deceiving and the public is already confused as it is. The gigaflop rating is a good idea but I’m not sure if it’s understood by the masses. Intel’s next generation processor, the Itanium, only runs at 800-900 Mhz. I’m curious how their marketing will change over the next year. The P4 was/is a flop, they’re slashing prices just to get them into boxes these days. They’re currently at 2Ghz but I think the public is getting just a bit smarter and realizing the very small increase in performance isn’t worth the purchase of an entirely new system.

People can only surf the web so fast,a DSL or Cable connection is a much better idea in terms of increasing your system’s usability than buying a machine with a faster processor. It’s a shame that numbers are percieved as performance but educating your friends and relatives can help. If you know someone who is in the market for a computer, offer up some info on a mac and show them the facts behind processor architecture and relative speed. I have converted many a friend to the mac side by showing them the facts, all of them made the decision for themselves.

Tim

In the pre-PPC time Apple didn’t tell what MHz their processors were, but then they made similar models with only a different processor and had no choice but to walk the MHz path (how else could you tell the machines apart?)
I would think it’s morally wrong to sell an AMD 1.4 GHz chip in a box with the text 1700 on the side, thereby implying that the processor runs at 1.7 GHz. I would not like the idea of Apple following that path. All of you who looked at the   have seen how Amazon set it up. No Gigahertzes to start with, just categories like family, business and multimedia. And then decide whether a computer fits in such a frame. It’s the same as Apple tries at the Apple Stores. Bringing you a solution to a problem. And in such situations there’s no need for a Gigaherz rating.

But then again I could also say I have TWO processors both running at 5.0x10^8 Hz
One question, is there a way to find out how many GigaFlops your processor has? I know the dual 800 does 11.8, but how about the other new G4s?
_________________
One thing tho, the person that finds her, gets to name her, right?
Then I wanna name her Microsoft, cause she’s a vicious live-sucking bitch from which there is no escape.
That’s sweet Carl

<font size=-1>[ This Message was edited by: hoytt on 2001-08-31 11:18 ]</font>

I don’t think Apple should hide their speed ratings. However, they should play up their strengths more. Ask my lover, the subject of speed is overplayed, she would rather I had a bigger hard drive and dual processors.

A sersiousness aside, I choose Macintosh for the operating system and elegance of the way things are done. To be sure, if my Macintosh ran too slow, I would have to rethink my position, but right now I have more than enough speed.

One question, is there a way to find out how many GigaFlops your processor has? I know the dual 800 does 11.8, but how about the other new G4s?

You know nothing of the sort, Apple CLAIM 11.8 GFLOPS in their marketing but what on Earth does that prove? If you would like to run some benchmark on your machine that will return a result in FLOPS, try AltiVec Fractal Carbon. The Appleseed people have some opinions on exactly what this app measures though…

Apple may as well hide the clock speed of their chips, as they apparently refuse to compile and run SPEC benchmarks on them so we can actually make some fair comparisons with other processor designs.

I think AMD is shooting themselves in the foot with this scheme. Imagine now the Athlon 1.4 Ghz named Athlon 1700 and Intel following with the same naming scheme calling their P4 2 Ghz the Pentium 2300. AMD would still lag behind in this number game. Clearly, I don’t think Apple should go this way at all. I think waiting is the best medicine here. As soon as Intel has to market their Itanium 64-bit processor everything looks different. Suddenly, Apple’s PPC chips are ahead of the game because by then they have the G5 at 1.2Ghz while Intel’s Itanium runs at 800Mhz. If you’ve listened to the Macworld keynote speech, Apple’s Vice-President of Hardware (forgot his name) explained the difference in chip design and why the P4 is actually much slower than the PPC with the same clock rate. The Itanium will have a much more similar architecture in terms of the amount of pipes in the chip as the PPC. Because of that Intel can’t run the Itanium at a speed like their P4. However, the G5 is much better designed with future generations in mind. The whole PPC architecture is fairly new and has the benefit of a design that was planned way ahead of time. Intel’s CISC model is running out of steam. What we are seeing here is Intel’s last attempt to win the race of Mhz. They’ll shoot themselves in the foot as well with this 20 pipe P4 processor that sucks energy like a toaster and heats up like an oven. I think Apple does already the right thing by telling how it is (stating their Mhz clock rate) and educating people about the Mhz myth as well as doing speed comparisons with the Intel’s processors. Eventually, Apple will surpass Intel in this game. Just wait and see.

I actually think it is a good idea to have some a different naming /identifying convention for how a processor performs. However I do NOT think that naming it what would be a comparable number for what Intel’s speeds really are is a good idea (ie. a 867 G4 is named the G4 1900. If you do this then you do in fact propagate the mhz myth AND you say that Intel is the standard by which to measure. And I cannot except that! IMHO the best solution is to come up with a different number other than mhz where performance and workload are measured and not clock cycles. Something that the industry can agree on. 

2Ñž

As soon as Intel has to market their Itanium 64-bit processor everything looks different. Suddenly, Apple’s PPC chips are ahead of the game because by then they have the G5 at 1.2Ghz while Intel’s Itanium runs at 800Mhz. If you’ve listened to the Macworld keynote speech, Apple’s Vice-President of Hardware (forgot his name) explained the difference in chip design and why the P4 is actually much slower than the PPC with the same clock rate. The Itanium will have a much more similar architecture in terms of the amount of pipes in the chip as the PPC. Because of that Intel can’t run the Itanium at a speed like their P4. However, the G5 is much better designed with future generations in mind. The whole PPC architecture is fairly new and has the benefit of a design that was planned way ahead of time. Intel’s CISC model is running out of steam. What we are seeing here is Intel’s last attempt to win the race of Mhz. They’ll shoot themselves in the foot as well with this 20 pipe P4 processor that sucks energy like a toaster and heats up like an oven.

I hardly know where to start debunking this…

Firstly, there’s no point in making claims about the G5 until it officially exists and is tested - all else is myth and rumour. If you’d like to look at some IA-64 results compared with other current processors, check out http://www.spec.org.

Secondly, if Apple were such IPC purists, why did they move from the MPC 7400 series to the 7450? Because whichever way you stack it up, the 7450 offers a lower IPC figure. MacSpeedZone have recently discovered to their dismay that the old 533Mhz G4 system can outperform the current 733Mhz G4 system. This is a BAD THING.

Thirdly, the P4 has a 20 STAGE pipeline, not 20 PIPES. Good grief.

Lastly, CISC? RISC? What? The P4, PIII, Athlon AND G4 are all considered “post-RISC” designs. The x86 chips all decode the x86 instructions fed to them and then execute “micro-ops” (Intel’s name) in a RISC back-end. The G4 has an extended instruction set that allows better use of it’s additional execution units. Like AltiVec.

Please take time to peruse the many facts available via the web before posting next time. It’s really tiring to read Apple’s marketing spiel regurgitated onto this forum time and time again.

The gigaflop rating is a good idea but I’m not sure if it’s understood by the masses.

But you do understand it, right?

The P4 was/is a flop, they’re slashing prices just to get them into boxes these days. They’re currently at 2Ghz but I think the public is getting just a bit smarter and realizing the very small increase in performance isn’t worth the purchase of an entirely new system.

I’m not sure that you could consider the P4 a flop - although the public probably have their bulls*** detectors on red alert in the current economic climate.

If you know someone who is in the market for a computer, offer up some info on a mac and show them the facts behind processor architecture and relative speed. I have converted many a friend to the mac side by showing them the facts, all of them made the decision for themselves.

I certainly agree with you that a Mac is a much more thoroughly considered “Personal Computer” than any box running Windows, but trying to claima performance lead for the Mac platform is completely without foundation.

[Edited for profanity. Please, ladies and gentlemen, do try to monitor your own language.—R]
_________________
“I have won the essay writing competition, of that there is NO doubt!” - Alan Gordon Partridge

<font size=-1>[ This Message was edited by: alanpartridge on 2001-08-31 12:35 ]</font>

<font size=-1>[ This Message was edited by: Ricky on 2001-09-03 08:57 ]</font>

————————————————————————————————————-Quote

>Thirdly, the P4 has a 20 STAGE pipeline, not 20 PIPES. Good grief.
————————————————————————————————————-

A little hard on the guy on the basis of terminolgy, don’t you think.  The point is there are more stages on pentiums.

————————————————————————————————————-
Quote

>Lastly, CISC? RISC? What? The P4, PIII, Athlon AND G4 are all >considered “post-RISC” designs. The x86 chips all decode the x86 >instructions fed to them and then execute “micro-ops” (Intel’s name) in a >RISC back-end. The G4 has an extended instruction set that allows better >use of it’s additional execution units. Like AltiVec.
————————————————————————————————————-

Yes, but pentiums and PPC started from very different paradigms of design.  The fact that pentiums originally took a more CISC approach while PPC took a RISC approach is worth noting.  Sure, each side has slowly adopted design approaches of the other, but in order to contain both processors in their present states, “post-RISC” would have to be so general a term as to be meaningless.

All that x86 decoding takes transistors.  No such process is necissary (or wanted) on a PPC. I’d speculate that Intel would axe the decoding if it weren’t necissary for backward compatibility.  And while its true that new execution units, like altivec, and instructions have appeared in the PPC architecture, there are still far more instructions in the x86 architecture.

While your assertion that x86 isn’t pure CISC and PPC isn’t pure RISC is valid, there are many crucial differences in the instruction set currently employed by each architecture.

———————————————————————————————————-
Quote
>>———————————————————————————————————-
>>Quote:
>>
>>The gigaflop rating is a good idea but I’m not sure if it’s understood
>>by the masses.
>>
>>———————————————————————————————————-
>
>
>But you do understand it, right?
————————————————————————————————————-

Could you be more condescending?  Honestly…

————————————————————————————————————-
Quote
>I certainly agree with you that a Mac is a much more thoroughly >considered “Personal Computer” than any box running Windows, but trying >to claima performance lead for the Mac platform is completely without >foundation. 
————————————————————————————————————-

You don’t need to claim a lead.  Why not just try to inform the public that MHz are not king and that a 500Mhz G4 can at the very least keep up with a P4 running at twice the MHz rating. I’m running a 400MHz G4 at home and I’m just as happy with its performance as with my 1 Gig P3 at work.  Sure its anecdotal, but I don’t think that otherwise uninformed members of the general public would consider the G4’s performance to be bad at all.

I think that somehow breaking up the myth of MHz completely determining the speed of systems would be a good thing.  Floating point operations per second (FlOPS) run into similar problems in that they measure only a subset of what goes into making a system “fast”. I do, however, think they are at least a bit better; at least they measure, to some degree, work actually accomplished and not how fast the clock on the processor ticks.

But then, what about integer operations per second? Hard disk speed?  Ability to multitask effectively via large amounts of fast RAM? Even efficiency and design of the OS or Window Manager or file manager(Windows vs. MacOS, Finder vs. Explorer)?  The list of other factors is large, making a good “system rating” a hard thing to put your finger on.

To sum up: not everything is black and white. Lets all try to remember that.

Matthew