Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

A megahertz question

Originally posted by Vlade
An example is having a cluster of processors at lower clock speeds, such as a big computer cluster and a REALLY fast single chip. Your logic would suggest that the cluster is inefficient because has LOTS of computers at slow speed, and therefore the single fast efficient CPU would be better because it is designed better (despite the fact that it is 2-4 times as slow).
Click to expand...
No, my logic says Speed isn't everything. That's the whole point I'm trying to make here.

The Pentium 4 is poorly designed. Intel just did it intentionally to make the chip faster. The result is that a 2 GHz P4 will be beaten out by a 1.5 GHz G4 or G5 (and probably a good AMD, too), but Intel will be able to make a 3 GHz P4 in order to compensate for that.

The P4 is poorly designed because it would be blown away by processors at the same speed. Intel knew this when they designed the chip. They also knew that they'd be able to increase the chip's speed quite a bit in a relatively short period of time, thus compensating for the poor design.

Of course, if the G5 does reach 3 GHz next year, Apple will really be giving Intel a run for its money (just imagine, a dual-3 GHz Powermac... :D ).
 
Originally posted by pyrotoaster
Of course, if the G5 does reach 3 GHz next year, Apple will really be giving Intel a run for its money (just imagine, a dual-3 GHz Powermac... :D ).
Click to expand...

Your forgetting that intel will be using 4GHZ or 5GHZ CPUs by then, so it won't be as big as we think, BUT it will still be 50% faster than the current powermacs, and it will tempt me to get that instead of a new powerbook G5 when it comes out.


My main point is you need to redefine your word poorly, if intel had a 10GHZ chip that was only a little faster than its 3GHZ chip right now, it WOULD NOT mean its poorly designed, it means intel is doing a better job than IBM and AMD, so their chip is superior, and better designed (they designed it to scale up so high in clock speeds).

My last post, I don't think you see my point
 
Originally posted by Vlade
Your forgetting that intel will be using 4GHZ or 5GHZ CPUs by then, so it won't be as big as we think, BUT it will still be 50% faster than the current powermacs, and it will tempt me to get that instead of a new powerbook G5 when it comes out.
Click to expand...
4 GHz, maybe. But even then, not in consumer machines. The Pentium 4 is reaching its limit. Unless Intel finds some new breakthrough in their design, a 5 GHz processor would probably burst into flames.
My main point is you need to redefine your word poorly, if intel had a 10GHZ chip that was only a little faster than its 3GHZ chip right now, it WOULD NOT mean its poorly designed, it means intel is doing a better job than IBM and AMD, so their chip is superior, and better designed (they designed it to scale up so high in clock speeds).
Click to expand...
You still don't get my point. If they had a 10 GHz chip that was only slightly faster than a 3 GHz chip, it would be very poorly designed. It might still be fast, but that in no way means that it's well designed.

You're confusing design and speed. Just because something's fast doesn't mean it's well designed. Similarly, something slow could be well designed. For instance, the Pentium 3 (or the G4, for that matter) is a well designed processor. It's not as fast as a Pentium 4, but that doesn't change the fact that it's well designed.

Intel made the P4 poorly designed so they could make it faster. It wouldn't be able to hold a candle to an IBM or AMD at the same speed, but Intel can make its clock speed high enough that this isn't a problem for Intel (yet). By next year, IBM and AMD will both be giving Intel tough competition.
My last post, I don't think you see my point
Click to expand...
I see your point, but you obviously don't see mine. It seems like you're selectively reading my posts and only responding to comments taken out of context. In your last post all you quote is the last sentence of my post. And you did the exact same thing in the post before that.

The point is, just because the Pentium 4 is fast it isn't necessarily "well designed". It's just fast. In fact, the P4 is poorly designed, because processors even 500 MHz slower than it would actually be faster than it.
 
Originally posted by pyrotoaster
The point is, just because the Pentium 4 is fast it isn't necessarily "well designed". It's just fast. In fact, the P4 is poorly designed, because processors even 500 MHz slower than it would actually be faster than it.
Click to expand...


I see your point, I just don't agree with it. I think the end result is what defines how poorly or good a product is. I don't give a damn if it is 1MHZ or 100GHZ, or anywhere in between, as long as its end result is lots of operations per second, and for the last year or 2 the Pentium 4 has lead in performance.


Originally posted by pyrotoaster
Intel made the P4 poorly designed so they could make it faster. It wouldn't be able to hold a candle to an IBM or AMD at the same speed, but Intel can make its clock speed high enough that this isn't a problem for Intel (yet). By next year, IBM and AMD will both be giving Intel tough competition.
Click to expand...

I think that it was well designed, because the designed it in a way that could make it scale up to such high clock speeds. It DOESN'T matter if a AMD or IBM beets intel at the same clock speed because THEY ARE NOT AT THE SAME CLOCK SPEED, intel designed the chip so good so this wouldn't matter.
 
To reinforce Pyrotoaters point, how about this. Here are some fp2000 scores.

Pentium 4 3.2Ghz 1285
XEON 3.2GHZ 1213
Opteron 2GHz 1339
SPARC III 1.2GHz 1344
Power4+ 1.7GHz 1699
Itanic 2 1.5GHz 2119
SPARC 1.35GHz 1340

So their rankings are as follows, Itanic 2, Power4+, SPARC III, SPARC (Fujitsu), Opteron, Pentium 4 and finally the XEON. So the two Pentium chips are at the bottom of the field.

Per GHz.
Pentium 4 3.2Ghz 401
XEON 3.2GHZ 379
Opteron 2GHz 669
SPARC III 1.2GHz 1120
Power4+ 1.7GHz 999
Itanic 2 1.5GHz 1412
SPARC 1.35GHz 992

So the efficiency per GHz is as follows: Itanic 2, SPARC III, Power4+, SPARC (Fujitsu), Opteron, Pentium 4 and then the XEON.
 
Originally posted by Lanbrown
So their rankings are as follows, Itanic 2, Power4+, SPARC III, SPARC (Fujitsu), Opteron, Pentium 4 and finally the XEON. So the two Pentium chips are at the bottom of the field.

and

So the efficiency per GHz is as follows: Itanic 2, SPARC III, Power4+, SPARC (Fujitsu), Opteron, Pentium 4 and then the XEON.
Click to expand...
Interesting. Thanks for posting those. :)

Vlade: This is just my point. I can understand if you don't agree with it. But you have to admit, the P4 (along with its cousin, the XEON) definitely isn't efficient. Speed is only one factor in a processor's real power.
 
Although I know it worked as a marketing gimmick, I wish AMD had not used artificial numbers for the Athlon. Their labeling scheme helped people to compare Athlons to Pentiums but it also caused people to forget about the real clock speed of the AMD chips. This only helped to reinforce the MHz myth.
 
Originally posted by pyrotoaster
Vlade: This is just my point. I can understand if you don't agree with it. But you have to admit, the P4 (along with its cousin, the XEON) definitely isn't efficient. Speed is only one factor in a processor's real power.
Click to expand...

Speed to me is defined as work per clock times MHZ, so the pentium 4 is a fast chip. Those other CPUs are VERY fast, I wish we could get them in desktops for under 10000 bucks :p
 
Originally posted by Lanbrown
To reinforce Pyrotoaters point, how about this. Here are some fp2000 scores.

Pentium 4 3.2Ghz 1285
XEON 3.2GHZ 1213
Opteron 2GHz 1339
SPARC III 1.2GHz 1344
Power4+ 1.7GHz 1699
Itanic 2 1.5GHz 2119
SPARC 1.35GHz 1340

So their rankings are as follows, Itanic 2, Power4+, SPARC III, SPARC (Fujitsu), Opteron, Pentium 4 and finally the XEON. So the two Pentium chips are at the bottom of the field.

Per GHz.
Pentium 4 3.2Ghz 401
XEON 3.2GHZ 379
Opteron 2GHz 669
SPARC III 1.2GHz 1120
Power4+ 1.7GHz 999
Itanic 2 1.5GHz 1412
SPARC 1.35GHz 992

So the efficiency per GHz is as follows: Itanic 2, SPARC III, Power4+, SPARC (Fujitsu), Opteron, Pentium 4 and then the XEON.
Click to expand...

Just in case you don't know, the Power 4+ isn't a G5. The G5 is a PowerPC 970 which is a stripped Power 4+. A G5 is a lot less efficient than an Athlon 64/Opteron, as one Athlon 64 2.2 GHz beats a dual G5 2.0 GHz system. I don't know where you found bechmarks for a 3.2 GHz Xeon as they don't exist yet (unless it was a Pentium 4 EE which is in every way faster than the Pentium 4 because it manages to use the same 800 MHz FSB, so it should have a higher score).
 
Do I sense a pattern with Jonathan Amend's posts? They all seem to be saying how great Intel, AMD and Microsoft are and how all of Apple's products suck. :rolleyes: Doesn't he have anything* good to say about Apple?
 
I noticed your orginal disclamer when you first posted it. I still haven't seen a single post with a specific positive thing to say about Apple or Macs.

btw- iJon and I seldom agree so that is strange indeed.
 
Originally posted by yamabushi
I noticed your orginal disclamer when you first posted it. I still haven't seen a single post with a specific positive thing to say about Apple or Macs.
Click to expand...

Nobody ever said I have to. I don't remember there being a giant warning during the registration process telling me that I have to post about how great Macs are all day. I never said anything against Macs, other than mentioning that they have both strengths and weaknesses, as do PCs.
 
So do you mean that you don't really care for Macs and are just here for the purpose of starting arguments?
 
Originally posted by Jonathan Amend
Just in case you don't know, the Power 4+ isn't a G5. The G5 is a PowerPC 970 which is a stripped Power 4+. A G5 is a lot less efficient than an Athlon 64/Opteron, as one Athlon 64 2.2 GHz beats a dual G5 2.0 GHz system. I don't know where you found bechmarks for a 3.2 GHz Xeon as they don't exist yet (unless it was a Pentium 4 EE which is in every way faster than the Pentium 4 because it manages to use the same 800 MHz FSB, so it should have a higher score).
Click to expand...

I am full aware the Power4+ is not a G5, I never said it was.

The 3.2GHz XEON is available. Maybe you need to do your homework before you speak. Go look at dell and their Precision Workstation 650. Here’s an excerpt:
"TECH SPECS
Processor
Intel® XeonTM processors up to 3.20GHz with 1MB L3 cache (Dual Capable) 533MHz front side bus"

So, Dell provided those figures and the 3.2GHz XEON does exist.

The XEON posted a lower score then the Pentium 4.
 
Originally posted by Vlade
Speed to me is defined as work per clock times MHZ, so the pentium 4 is a fast chip. Those other CPUs are VERY fast, I wish we could get them in desktops for under 10000 bucks :p
Click to expand...

Is that some new fuzzy math? The P4 and XEON were at the bottom of the list.

You can get the others in a system for under $10,000. Most be that new math again.
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back