Canadian Mac Forums at ehMac banner
1 - 7 of 7 Posts

·
Registered
Joined
·
1,060 Posts
<BLOCKQUOTE>quote:</font><HR>Originally posted by Tomac:
http://www.adobe.com/motion/pcpreferred.html

We know it. We don't like hearing it, but we love our Macs anyway!
<HR></BLOCKQUOTE>

The first thing to take into account is that the tests performed were a P4 system with 2.5x the GHz power of the G4 system it competed against. Coupled with a series of faster components removed some of the bottlenecking that the G4 could use to it's advantage before.

Second is a new technology called Hyperthreading. This allows the processor to act almost like 2 processors. In effect, the numbers they show is that the P4 2.53GHz processor was performing like a 2 or 2.5GHz G4 processor. It's more efficient a way of doing things.

Is the P4 faster? Inherently no, in an even GHz to GHz comparison. But stack the deck in their favour (faster processor numbers, more streamlined internal bus with lower bottleneck, Faster RAM) and you get better results than the top of the line Mac.

Certainly the way they presented their findings makes the Mac look sluggish in comparison, but it's not exactly the fairest of comparisons, since the G4 has been growing at a slower rate than the P4 and their internal bus has been better optimized than previous machines.

Now I'd like to see how a 2.53 or 3.06 GHz P4 stacks up against a 2.53 or even 3.06GHz G4. Even without hyping and tweaking the internals to match, the G4 will squeak ahead of the P4.

Naturally, the PC marketing monkeys will have a field day with stuff like this...
 

·
Registered
Joined
·
2,198 Posts
Strongblade wrote:
Second is a new technology called Hyperthreading. This allows the processor to act almost like 2 processors.

HyperThreading is an interesting bit of technology, but I think people have over-estimated the performance win it brings. A HT-enabled processor does not behave like two (otherwise equivalent) non-HT-enabled processors; our own internal tests show the speedup to be 20-30% compared (where dual processors would provide an 80+% speedup).

Now I'd like to see how a 2.53 or 3.06 GHz P4 stacks up against a 2.53 or even 3.06GHz G4. Even without hyping and tweaking the internals to match, the G4 will squeak ahead of the P4.

The problem is that Motorola can't produce a 2.53GHz (let alone a 3.06GHz) G4, while Intel can produce a 3.06GHz P4. The G4 might have a nicer architecture, but if Motorola can't drive it at speeds high enough to compete with the P4, then it's kind of a moot point.
 

·
Registered
Joined
·
1,060 Posts
<BLOCKQUOTE>quote:</font><HR>Originally posted by jfpoole:
Strongblade wrote:
Second is a new technology called Hyperthreading. This allows the processor to act almost like 2 processors.

HyperThreading is an interesting bit of technology, but I think people have over-estimated the performance win it brings. A HT-enabled processor does not behave like two (otherwise equivalent) non-HT-enabled processors; our own internal tests show the speedup to be 20-30% compared (where dual processors would provide an 80+% speedup).

Now I'd like to see how a 2.53 or 3.06 GHz P4 stacks up against a 2.53 or even 3.06GHz G4. Even without hyping and tweaking the internals to match, the G4 will squeak ahead of the P4.

The problem is that Motorola can't produce a 2.53GHz (let alone a 3.06GHz) G4, while Intel can produce a 3.06GHz P4. The G4 might have a nicer architecture, but if Motorola can't drive it at speeds high enough to compete with the P4, then it's kind of a moot point.
<HR></BLOCKQUOTE>

True, I overstated the Hyperthreading ability, but coupled with the higher raw processor speed and other internals, it shows why it gets a better score.

As for the comparison fo raw vs raw being moot? I don't know about that. Just because one doesn't exist doesn't make the point invalid.

:cool:
 

·
Registered
Joined
·
57 Posts
strongblade - yes, mhz per mhz, the g4 is 'faster' than a p4. However, the fact that you cant get a faster g4 at the moment coupled with the fact that they are relatively similar in cost makes it a fair comparison in my opinion.
 

·
Registered
Joined
·
259 Posts
Discussion Starter · #6 ·
Motorola's basically dead in terms of furthering Apple -- stand still, not much innovation or care for getting chips out when Apple needs 'em. Example: 1.42 GHz G4s. Apple doesn't have a chance if they continue to rely on them for chips.

It's up to IBM for the NEXTSTEP (heh) in furthering Apples quest for RAW horsepower.
 

·
Premium Member
Joined
·
5,247 Posts
" ... HyperThreading is an interesting bit of technology, but I think people have over-estimated the performance win it brings. ..."

In fact, HT actually slows down some operations. There are many Windows apps that perform faster with HT disabled, and some applications require it be disabled in the BIOS to run at all.
 
1 - 7 of 7 Posts
Top