Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Status
Not open for further replies.
Originally posted by centauratlas
>what IS in those purported boxes at the Apple stores that said<

G5 demos

>why did the VeriTest comparisons not have any AMD processors on it?<

Because from a marketing standpoint the public knows "Intel inside" and there was no reason to confuse the issue. Plus I think they were differentiating between what is in the same market space (e.g. desktop processors).

>G5 iMac

12-18 months at the earliest. Just a guess though


>FibreChannel port on that new Powermac and I'll agree with you.

Just look at the Apple web site in the store under a custom option. Pick "Fibre Channel Card".
There's no way those boxes contained G5 demo machines. Apple Retail Stores won't be receiving shipments of the new Powermac until September.

The G5 iMac may arrive as soon as MWSF04. I'd say a 6 to 12 month timeframe is more likely.

Also, you obviously didn't read the MacBidouille rumor. It didn't say there was a Fiber Channel card. In fact, that would've made no sense because they claimed to be looking at a Motherboard. MacBidouille clearly claimed that there were two Fibre Channel ports on the Motherboard.
 
Re: Why oh why can't we all just join hands and sing of rainbows and bunnies...?

And with regards to Macmax's original post:

I think you're being a tad harsh on the bloke (?). After all, he just seems to be pointing out that Apple chose a set of test results that showed off the G5 in the best possible light.


what i am saying is the choice of words he had, strange.

And , yes the opteron might be faster, but as they said, it is not yet in a personal computer , but in servers and powerstations, so no one seems to be a liar.
Anyway, it was a great day for the Mac
 
Originally posted by rog
I seriously doubt that. If we're lucky, the DP 2 GHz will be in stores in early September, maybe in volume a few weeks after that. They aren't going to ramp up that quickly. Some are saying the 90nm process will be needed to get beyond 2GHz. 3GHz in a year is not that impressive. 50% faster in 12 months is not even keeping up with the distorted "moore's laws" view which says 100% faster every 18 months (yes, I know the real moore's law is 100% more transistors every 18 months, but Wintel has been able to basically meet the speed based definition). I know the P4 is not SMP capable, but with a 3.2 out, that means a Xeon is not far behind at that speed, if not out already.
Check you date math. In Jan '03 the Mac was at 1.42GHz, in ~July '04 the Mac will be at 3GHz; that's 2x speed increase in 18 months (yes, there was a processor change in that time so the comparrison isn't totally valid but its not totally invalid either).
 
Estimations and analysis part 1

Some estimations for the newer Pentium 4s as well as an analysis of the compilers used:

Submitter: IBM
CPU: PPC 970 1.8 GHz
Compiler: Visual Age?
SPECint Base: 937
SPECfp Base: 1051

Submitter: Intel
CPU: Pentium 4b 2.80 GHz
Compiler: ICC 7.0
SPECint Base: 976
SPECfp Base: 915

1) Pentium 4 2.8 GHz is 4% faster than PPC970 1.8 GHz on SPECint

2) PPC970 1.8 GHz is 14% faster than Pentium 4 2.8 GHz on SPECfp

Submitter: Apple
CPU: Pentium 4b 3.06 GHz
Compiler: GCC 3.3
SPECint Base: 889
SPECfp Base: 693

Submitter: Apple
CPU: PPC970 2.0 GHz
Compiler: GCC 3.3
SPECint Base: 800
SPECfp Base: 840

1) Pentium 4 3.06 GHz is 11% faster than PPC970 2.0 GHz on SPECint

2) PPC970 2.0 GHz is 21% faster than the Pentium 4 3.06 GHz on SPECfp

From this comparison we can derive that the single CPU SPEC scores match well with the original SPEC marks done by IBM and are therefore, most likely accurate.
 
Estimations and Analysis: Part 2

Now let’s look at how some of the faster Pentium 4’s would fare against the 2 GHz PowerPC 970.

Submitter: Intel
CPU: Pentium 4b 3.06 GHz
Compiler: ICC 7.0
SPECint Base: 1099
SPECfp Base: 1077

Submitter: Intel
CPU: Pentium 4c 3.00 GHz (800 MHz FSB)
Compiler: ICC 7.0
SPECint Base: 1164
SPECfp Base: 1213

1) The Pentium 4c 3.00 GHz is 6% faster than the Pentium 4b 3.06 GHz in SPECint

2) The Pentium 4c 3.00 GHz is 12% faster than the Pentium 4b 3.06 GHz in SPECfp.

Submitter: Intel
CPU: Pentium 4b 3.06 GHz
Compiler: ICC 7.0
SPECint Base: 1099
SPECfp Base: 1077

Submitter: Intel
CPU: Pentium 4c 3.20 GHz (800 MHz FSB)
Compiler: ICC 7.0
SPECint Base: 1221
SPECfp Base: 1252

1) The Pentium 4c 3.20 GHz is 11% faster than the Pentium 4b 3.06 GHz in SPECint

2) The Pentium 4c 3.20 GHz is 16% faster than the Pentium 4b 3.06 GHz in SPECfp.
 
Estimations and Analysis: Part 3

So calculating these increases over the Pentium 4 3.06 using the GCC compiler, we would get the following:

Submitter: Estimation
CPU: Pentium 4c 3.00 GHz (800 MHz FSB)
Compiler: GCC 3.3
SPECint Base: 942
SPECfp Base: 776

Submitter: Estimation
CPU: Pentium 4c 3.20 GHz (800 MHz FSB)
Compiler: GCC 3.3
SPECint Base: 987
SPECfp Base: 804

Now we can compare these scores to those of the 2 GHz PPC970 as well as the older 3.06 GHz Pentium 4b

Submitter: Apple
CPU: Pentium 4b 3.06 GHz
Compiler: GCC 3.3
SPECint Base: 889
SPECfp Base: 693

Submitter: Apple
CPU: PPC970 2.0 GHz
Compiler: GCC 3.3
SPECint Base: 800
SPECfp Base: 840

Comparison between 3.20 GHz Pentium 4c and 2.0 GHz PPC970:

1) 3.20 GHz Pentium 4c (800 MHz FSB) is 23% faster than 2.0 GHz PPC970 on SPECint

2) 2.0 GHz PPC970 is 4% faster than 3.20 GHz Pentium 4c (800 MHz FSB) on SPECfp

Comparison between 3.00 GHz Pentium 4c and 2.0 GHz PPC970:

1) 3.00 GHz Pentium 4c (800 MHz FSB) is 18% faster than 2.0 GHz PPC970 on SPECint

2) 2.0 GHz PPC970 is 8% faster than 3.00 GHz Pentium 4c (800 MHz FSB) on SPECfp

So at 2.0 GHz, the PPC970 still compares rather well against the newer Pentium 4s, against a 3.2 GHz Pentium 4C, it is 23% slower in SPECint but 4% faster in SPECfp, similarly, against a 3.0 GHz Pentium 4C, it is 18% slower in SPECint but 8% faster in SPECfp.
 
Re: why 2nd optical drive?

Originally posted by Chomolungma
Are you burning DVD/CD 24 hours a day seven days a week? Obviously, it is hard to see any of us doing this, unless you are trying to burn prirated software in Thailand. A small porn business in home?:D Just teasing.

Personally, I don't see a need for a CD-RW or a superdrive. With hard disk back up solution and iPod, optical disk seems so passe:D . Large file that need to be deliver, can be done using P2P with broadband or ftp.


I would like to have the option to install a faster internal cd-rw drive. The super drive is nice but a 52x burner can be had for under $100. It will cost twice as much for an external...plus it takes up more desk space. This has been a complaint for many people for a while now. It is just dissapointing to finaly include it in one revision and they take it away in the next. It is not just about transfering files or data storage but making music and dvd backups in the least amount of time possible. I think the whole point of making computers faster is to get to a point where we do not wait for computers any more....they wait for us. I have a 2x superdrive that burns cd-r's at 8x. But I do not belive I should be limited to that speed. I don't want to wait 10 minutes when I could be done in less than 2. Any way my point is offer the expandability.
 
SO, Tech-People, I have a riddle...

In the "REAL WORLD" which is really faster/better. My SCSI 160 drives, or these new Serial ATA drives?

Don't flame me, just inform me...
 
Don't Worry, Be Happy.

I've been working on a G4-450, at home and at work, since 2000 and I don't know what it feels like to work on anything faster. These machines are going to ROCK!!! (Hell, a G4 1.42 would rock right now.)

I don't need more than 3 PCI slots, 2 optical drives or more hard drive bays. If I need expansion, I have Firewire and LaCie's d2 (Aluminum) line of drives for Optical and Hard drives.

The only thing I will miss is the side door. Though I didn't upgrade much, I loved popping that open and blowing my PC using friend's minds with the ease of access.

Also for everyone jumping on the comparisons beetween the G5 and Xeon, P4 and Opteron, you still have to take into account real world performance while using machines. How hast they FEEL! With Panther, G5 and a bottleneck free system, these systems are going to feel faster than any bloated PC running winblows.

The spinning beach ball is dead.


iN8
 
Originally posted by ddtlm
Steradian:

Don't get too excited about SPEC scores. Notice Apple did not invite an Opteron? :) Thats cause Apple would have lost.


That is not the reason.

1. The Opteron is not widely available. Try to buy a system from Dell with a Opteron.
2. Most but not all of the Opteron's speed increase requires a 64bit optimized OS. The last I checked the 64bit version of win2k3 server has not been released.
3. The Opteron is a server processor. I know that the Xeon is also. But the Xeon can be bought in stock desktop system from large computer manufactures (Dell). The Opteron cannot. It can only be had in actaul servers.

I am a windows user/developer. I am also an AMD fan. The Athlon is one of the best processors ever designed. It really made Intel scramble. But, the Opteron is not the second comming. It should be good, but the next P4 core design that Intel will bring out later this year should ramp over the 4Ghz mark. It will take a powerful x86 design running at < 2Ghz to match the processing speed there. I don't know if it will. (The G5 has the benefit of not being x86.) Besides, the consumer version (Athlon 64) will not be out until later this year. That would be the valid comparison.

Oh, I should probably clarify, I bought a mac as a second home system for kicks because I thought that idea of a Unix based mac was pretty cool and if I didn't like it, I could always put a ppc linux on it. But I did like it. Now I mainly, but not completely, use my mac for normal home use. I really like the mac and my next laptop will probably be a powerbook. But I will never completely switch because I still need my Windows box to code on. (I work for a windows only shop and VPC sucks bad compaired to the real thing.)
 
Shipments of the PowerMacs delayed

Spymac is now reporting that the new PowerMacs are being delayed on or before September 2.

Either this is an indication of the release of Panther or they pulling another "Apple premature notification" like they did with the 17" Powerbooks.

Good thing I didn't order one yet. I would be mad as hell.

CheekyGit :cool:
 
Re: Re: why 2nd optical drive?

Originally posted by User X
I would like to have the option to install a faster internal cd-rw drive. The super drive is nice but a 52x burner can be had for under $100. It will cost twice as much for an external...plus it takes up more desk space. This has been a complaint for many people for a while now. It is just dissapointing to finaly include it in one revision and they take it away in the next. It is not just about transfering files or data storage but making music and dvd backups in the least amount of time possible. I think the whole point of making computers faster is to get to a point where we do not wait for computers any more....they wait for us. I have a 2x superdrive that burns cd-r's at 8x. But I do not belive I should be limited to that speed. I don't want to wait 10 minutes when I could be done in less than 2. Any way my point is offer the expandability.

I do believe the new G5 at least the lower end model can be confirgured to have a combo drive (please correct me if i'm wrong). Combo drive offers a faster CD burn rate minus the DVD-r feature off course. I think you made a good point with regard to having variable speed option!
 
Re: A Step Backwards in Expandability

Originally posted by mustang_dvs
I really could care less about the looks of the G5 (though I question the ergonomics of getting rid of the mobo hinge)

You're thinking with last year's specs. The machines come with 80, 160, or 250 serial ATA drives on their own channel. That means you'll have a real RAID system possible without havign to buy a third party interface. Also, hard drive sizes have got enormously big as of late... do you really need either 500 megs or a 250 raid in your box? Isn't anything in excess of that a job for a dedicated RAID system?

what bothers me is the apparently cramped confines of the interior, which limits us to 1 optical drive and two hard drives

This seems like a huge _improvement_ to me. Not having all the wires, cards, and associated plugs move when I swing open the door is a god send. It was great for accessability, but only if you had enough space around your machine to do it. I like the way it's done now.
 
Re: SO, Tech-People, I have a riddle...

Originally posted by Osglith
In the "REAL WORLD" which is really faster/better. My SCSI 160 drives, or these new Serial ATA drives?

Don't flame me, just inform me...

I have heard the academics basically say they are so comparable as to be almost the same. The only difference, of course, being the much cheaper price of anything that's missing the "SCSI" label. :)
 
Re: Re: i think this guy is a liar

Originally posted by MacBandit
Doesn't really matter now does it. All you have to do is watch the keynote and see the Dual G5 not just beat but completely pulverize a Dual/Xeon. So in the real world the Mac is king.

True dat! also, even if the guy was right and in the tests the xeon etc beat the G5, he didn't mention that those computers are over $1000 more expensive than the fastest G5. Well, in a way he might be right, but the truth is, we did see how slow the PC was in loading those layers of the poster..

I wan't more information about the machines and I think there will be loads of tests made when the computers come out. I'm not gonna get one for years and I'm actually hoping the chip to hit the laptops, but I do wanna know the truth - after all, the apps used in the demos were chosen by apple, not intel. then again, those apps were the ones that the buyers of a 2x2GHz most likely use..
 
plumbercrackboy:

You don't need Dell, Windows or even a workstation to run SPEC scores. Opteron workstation motherboards are starting to become available, and apparently whole machines are available from small (professional) computer makers such as boxxtech. There a lots of reasons Apple could provide for not including Opterons, but ignoring them for the reasons you provide is silly. Opterons are going to be one of the G5's biggest compeditors because they are also a foundation for affordable 64-bit workstations. Most places need Windows first I imagine, so that gives Apple some room to breathe, but still I think the Opteron is the big compeditor to the G5 once they both get onto the market with their respective mainstream 64-bit OS's. Who cares if Dell sells one, people who need 64 bits are going to get either G5's or Opterons where they are available.

In addition to making Apple look bad, the presence of another 64-bit platform would dampen their marketing a lot. So the Opteron was ignored.
 
Originally posted by venture160
i dont buy that its the fastest pc...... even though i love the G5 and i want one really badly... if you look at the benchmark test it compares it to a 3 ghz 533mhz bus dell... but it left out the 800hmz bus version that dell released a few months ago.... which is prob just as fast, not to mention not as expensive. I think apple needs to lower their prices to be competitive.

They compared it to the 533MHzFSB Xeon and to an 800MHz FSB P4.

533MHz is as fast as you can get on a Xeon AFAIK.

If you want dual processors on Intel, you ain't getting 800MHz FSB. Period.
 
Originally posted by rog
And a 3.2 GHz P4 was released today. By December, they'll be at 3.6 or so, while the G5 will still be at 2GHZ. Closer, but until we see more independent benchmarks from running real programs, there's no way to know who's really ahead. What they need to do is get some dualies on the low end. There's no reason now why Macs shouldn't win on performance AND price, especially since the G5 costs them less than the G4.

Ahhh, but here's the kicker:

In December Intel will be at 3.6GHz. That's a 20% improvement over yesterday's proc (although the FSB hasn't and won't be increasing unless Intel is going secret-ops on us). In June of next year, it won't be much higher (although the P5 will be the primary processor maybe, still running at 3.6GHz, but 10-15% faster per clock cycle than the P4). Assuming a 3.6GHz P5, that's an overall increase in performance of (drumroll ...) 44%, although likely none of it on the FSB.

In the same time period, Apple/IBM say they will have the 3GHz 970s out, for an increase from today's performance of a full 50%. And the FSB appears to scale directly with processor speed, which would put a 3GHz proc or two each with 1.5GHz of FSB.

The gap is widening long-term. Yes, there will be days where it is smaller than others, but in general ... life is good for Apple.
 
Status
Not open for further replies.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.