WWDC Coming...

[Frobozz]

Originally posted by boxcar
you would think they wouldn't be having a session on the G4 if the ppc970 was coming out.

... except for the fact that the Velocity engine's 162 instructions are identical on both the 970 and the G4. It doesn't matter if they put G4 or 970 in there, really, the 970 just does it faster. I'm referencing the recent ArsTechnica article concerning the AltiVec support in the 970.

 

[Vonnie]

Originally posted by jcr

iDVD, DVD Studio Pro, Apple's Project Builder, and the Apple Store are all examples of "complex" apps based on Objective-C. I personally have worked on derivatives-trading apps for a number of wall street firms that were written in Objective-C, and we've never wished we had the inflexibility that comes from so-called "type safety."

-jcr

And the other thing is, that you can get static type binding if you so desire. You can always do a NSString* myString instead of id myString.

But the added flexibility is a serious advantage, not a disadvantage, also in big projects. It requires a bit more discipline, but I don't have to start writing interfaces/protocols, or create weird inheritance hierarchies to do basic polymorphism.

 

[mathiasr]

Re: No "New" PowerMacs Coming At WWDC

Originally posted by pkradd
Apple will not introduce new PowerMacs at WWDC. They are history. A completely new form factor and NAME will be unveiled along with the preview of "Panther".

Do you mean "Xstation" based on "Xserve"?
A name change alongside the hardware overhaul sounds logical.

 

[Bengt77]

Re: Re: Re: No "New" PowerMacs Coming At WWDC

Originally posted by mathiasr
Do you mean "Xstation" based on "Xserve"?
A name change alongside the hardware overhaul sounds logical.

No way it'll be Xstation. Where's the iStation then? The 'i' vs 'Power' naming scheme has always existed. (Well, at least since July 1998, with the introduction of the original iMac.) So it'll never be called the Xstation, for there's no iStation. There is an iMac though, so if Apple decides to take over the 'x' from the Xserve for it's entire Pro line of computers, the Power Mac will probably be renamed the xMac.

I never liked the capital 'X' and lowercase 'serve' name-parts of the Xserve; think they should have named it xServe. Then, the xMac name would make sense (note that it sounds like Axe Mac: the Mac that'll finally take good care of all PCs). They could even rename the PowerBook (indeed very strange Power Mac is two words, while PowerBook is only one) the xBook. Sounds pretty cool, don't it?!

 

[job]

Re: How the hell??

Originally posted by deepkid
Let's keep an open mind and see what WWDC actually brings.

I'm beginning to wonder what will happen if the 970s are not released at the WWDC.

Mass riots?

 

[blueBomber]

Re: Re: How the hell??

Originally posted by job
I'm beginning to wonder what will happen if the 970s are not released at the WWDC.

Mass riots?

I imagine the conference hall will look something like the prison at the end of the movie Natural Born Killers

 

[BaghdadBob]

xMac sounds too much like x-Pac. Who is x-Pac? I don't remember...something about rapping. Not that I haven't thought of that myself.

How about PMX, as in PowerMacX...
"P-M-X!
"You thinkin' what if we got it, Billy we wid it, all you gotta do is buy it baby!
"Mac or die!
"What y'all PCs want...*unh, unh*...What y'all PCs want..."

 

[Snowy_River]

Re: Time will tell...

Originally posted by patmcfar8
...
Is anybody else bothered by those on forums that overhype every Apple event with unrealistic expectations based on rumors and wild speculation, and then pronounce their extreme dissapointment along with the "end of Apple" upon the passing of the event that hasn't lived up to their every hope and dream?

Sometimes I think us crazy Mac fans are our own worst enemies.

Having said all of that, I'd love to be working on a dual 2ghz 970 runnin' 10.3 by the end of this summer. Here's to eternal optimism...

Yes, those people bother me to no end. We must live in the real world, and while it's fun to hope and dream of the uberMac, if Apple isn't putting it out there is a good reason. Getting mad at Apple won't change that. Apple, after all, isn't really looking to screw us. They aren't trying to put out slower computers than the PC world...

Having said that, I'd love to be working on a new 1.6 GHz 970 15" PowerBook running 10.3 by the end of this summer. To eternal optimism...


<What?!? The new 15" PowerBook only has a 1.4GHz 970 in it?! Man that just sucks!! Apple is always short changing us! They could at least cut us a break and sell it at half the price of the old 15" TiBook!>

 

[Snowy_River]

Re: New Mac Stuff

Originally posted by painandgreed
...
Multiple people, while trying not to violate NDA, have informed me that I should believe the rumors...

Which rumors, though? The rumors are all over the place. That the 970 is real? That's really not a rumor, as IBM has pubicly announced the existance of this chip. That the 970 is coming this year? Well, it really doesn't take a rocket scientist to deduce that. That the 970 is coming for WWDC? Or that it's not coming for WWDC? There are rumors both ways. And what about PowerBooks? There are rumors both ways there, too.

Saying 'believe the rumors' really isn't enough information.

New PDA? I think we've already seen it. I think the iPod is our Apple PDA. With the next upgrade, I suspect we'll see feature bloat to include text editing, some more programs, and some neat dockable add ons that will make it comparable to other PDAs. They'll do this slow, keeping the iPod as a successful music player first and foremost, so if it flops as a PDA, they aren't saddled with another Newton (and without worring about getting enough of the PDA market away from other more established companies).

While you may be right about the iPod being, effectively, a new PDA, I just thought I'd offer a reality check. The Newton was not a flop. The plug was pulled years ago, and yet Newtons are still sold today. New software is still written today. New hardware expansion are designed (and sometimes even marketed and sold, rather than just being shared gratis). So, I think it's a bit of a stretch to say that the Newton was a flop.

 

[Snowy_River]

Re: How the hell??

Originally posted by deepkid
How the HELL do you people draw these sorts of conclusions???

That is about as ludicrous as saying "hey, he LOOKS guilty and everybody seems to think he is, so HE IS".

Until Apple makes an announcement, its all speculation.

Let's keep an open mind and see what WWDC actually brings.

While I don't agree with the statement that it is 'assured', it is a good bit of perspective. Apple is really pushing the WWDC this year. That does suggest that they have something big up their sleeves. And it does seem reasonable that that something would be the 970. Not assured, but reasonable.

However, I'm not going to hold me breath...
(Yes, I admit it, I'm addicted to air. Really, I'd try to quit, but the withdrawls are killer... )

 

[Snowy_River]

Re: No "New" PowerMacs Coming At WWDC

Originally posted by pkradd
Apple will not introduce new PowerMacs at WWDC. They are history. A completely new form factor and NAME will be unveiled along with the preview of "Panther". We'll see a bunch of new software apps from third party vendors (including that long-awaited and delayed one!) and the first non-beta version of Safari.

You seem awefully sure of yourself. I'm not saying that it couldn't be true, but we may have Power Macs around for a while longer...

 

[Snowy_River]

Originally posted by BaghdadBob
I doubt that, actually. The G3 is coming along and Apple doesn't want to deal with Moto anymore. I have predicted that Apple will sue out of their G4 obligations with the contractual failure by Moto with the G5, and we will see an all G3/970 line as soon as "Gobi" or whatever becomes a reasonable low-end processor.

Bye-bye Moto!

Well, even if Apple gives Moto the boot, I'd expect the IBM 750GX processor (G3 + Altivec compatibility) to be called a G4. Maybe the G4i, as it comes from IBM and would be used in the 'i' line? (i.e. no more G4 based Power Macs or PowerBooks...)

 

[Snowy_River]

Re: Re: Re: Re: No "New" PowerMacs Coming At WWDC

Originally posted by Bengt77
No way it'll be Xstation. Where's the iStation then? The 'i' vs 'Power' naming scheme has always existed. (Well, at least since July 1998, with the introduction of the original iMac.) So it'll never be called the Xstation, for there's no iStation. There is an iMac though, so if Apple decides to take over the 'x' from the Xserve for it's entire Pro line of computers, the Power Mac will probably be renamed the xMac.

I never liked the capital 'X' and lowercase 'serve' name-parts of the Xserve; think they should have named it xServe. Then, the xMac name would make sense (note that it sounds like Axe Mac: the Mac that'll finally take good care of all PCs). They could even rename the PowerBook (indeed very strange Power Mac is two words, while PowerBook is only one) the xBook. Sounds pretty cool, don't it?!

Actually, the 'Power Mac' terminology comes from the advent of the PowerPC chip. The new chips gave rise to the new line of 'Power Macintosh' computers. Now, the PowerBook is another matter, as it had that name before the 603e chip (the first PowerPC chip that was mobile-friendly) was available. The only mobile Mac that didn't bear the name PowerBook (until the iBook came along) was the very first 'Macintosh Portable'.

Indeed, this may well explain why 'Power Mac' is two words, and 'PowerBook' is one. It comes from their different heritage. It is only in more recent times that we have come to think of there being a 'Power' line and a consumer or 'i' line. While there have always been Macs aimed at the power users and Macs aimed at the consumers, we've only had the nameing conventions for a few years now.

 

[BaghdadBob]

Originally posted by Snowy_River
Well, even if Apple gives Moto the boot, I'd expect the IBM 750GX processor (G3 + Altivec compatibility) to be called a G4. Maybe the G4i, as it comes from IBM and would be used in the 'i' line? (i.e. no more G4 based Power Macs or PowerBooks...)

Hmmmm...interesting theory. Wouldn't there be some copyright restrictions? Maybe it'll be good enough to be the real G5. But, from a markting standpoint that really makes sense. I guess I won't give you crap for making a different post for all five of your replies because you make pretty good points

Fission is the future

 

[blueBomber]

that is a good theory... probably one of the most sensible ones heard in this thread yet

 

[fpnc]

PPC970 at WWDC 2004?

Frankly, I think we'll see PPC970-based Macs shipping at a time that is almost closer to NEXT year's WWDC than it will be to June 2003. That's my totally controversial leadoff statement.

IMO, it will be a truly amazing feat if Apple ships the PPC970 before the very end of this summer let alone June or July 2003. Late fall or early winter seems very likely to me and it's possible that we will still be waiting come January 2004.

As for the various rumors over the last few weeks, I think that many of these have been largely debunked. In fact, the recent Arstechnica paper on the PPC970 should have placed the last nail in the coffin on those rumored performance benchmarks.

As far as configurations, I expect that we will see SINGLE processor 970's replacing the current price points, with perhaps a more expensive "ultra" option with dual-970's (and Xserves). In the meantime we will likely see a speed bump in the G4 desktops, followed by a price reduction later in the year. And, of course, we're going to see a new and improved 15" PowerBook (completely new form factor).

When the 970's arrive they should be very nice machines but they won't be setting any new price-performance marks in the PC industry. They certainly will represent a new and vastly improved phase for the Macintosh but I won't be expecting miracles (such as dual 970's at less than $2K or single-processor machines that CLEARLY outperform high-end P4-based systems on a wide range of tasks).

 

[Snowy_River]

Originally posted by BaghdadBob
Hmmmm...interesting theory. Wouldn't there be some copyright restrictions? Maybe it'll be good enough to be the real G5. But, from a markting standpoint that really makes sense. I guess I won't give you crap for making a different post for all five of your replies because you make pretty good points

Fission is the future

Well, what can I say? I post replies as I read through the thread. Just because they all happen to pile up at what is the end of the thread when I'm reading it...

Thanks for not giving me crap about it...

 

[Snowy_River]

Re: PPC970 at WWDC 2004?

Originally posted by fpnc
...
IMO, it will be a truly amazing feat if Apple ships the PPC970 before the very end of this summer let alone June or July 2003. Late fall or early winter seems very likely to me and it's possible that we will still be waiting come January 2004.

I think that what's missing here is the impact that the chip producer has on the timeline. If the rumors are true and the 970 has had far fewer problems in production than anticipated, then they could well have production quality 970s shipping to Apple already. And please don't say that Apple needs time design the systems around these chips. Apple has probably been designing those systems since before the 970 went into final sampling. All they're doing once they get the production chips is fine tuning the systems to address any last minute bugs. And Apple can speed a process like that up, at least a little, if there is sufficient impetus to do so.

As for the various rumors over the last few weeks, I think that many of these have been largely debunked. In fact, the recent Arstechnica paper on the PPC970 should have placed the last nail in the coffin on those rumored performance benchmarks.
...

Yes, but the benchmark rumors are a far cry from being the only rumors regarding the 970 that have been floating around. To dismiss them all just because the benchmarks are questionable is dubious...

...
When the 970's arrive they should be very nice machines but they won't be setting any new price-performance marks in the PC industry. They certainly will represent a new and vastly improved phase for the Macintosh but I won't be expecting miracles (such as dual 970's at less than $2K or single-processor machines that CLEARLY outperform high-end P4-based systems on a wide range of tasks).

I think that it really comes down to the fact that we really don't know what the price-performance marks will be like. We could well be in for a time when we will recapture the 'Golden Age' of the Power Mac. Remember 'The G3 in every Power Mac is up to twice as fast as the Pentium', or 'Here at Apple, we'd like to appologize for publicly toasting the Pentium'? Ah, those were the days. I'm still hopeful that the 970 could bring back some aspects of those glory days.

 

[GregA]

Originally posted by fpnc
When the 970's arrive they should be very nice machines but they won't be setting any new price-performance marks in the PC industry. They certainly will represent a new and vastly improved phase for the Macintosh but I won't be expecting miracles (such as dual 970's at less than $2K or single-processor machines that CLEARLY outperform high-end P4-based systems on a wide range of tasks).

Yes ArsTechnica shows that the 970 will not be up to the P4 speed. Except for number crunching/Altivec(Velocity Engine).

Originally posted by Shaktai
The workshop mentioned is regarding the utilization of Altivec on the G4. A great many current developers are not fully familiar with Altivec and how it can be used to speed up performance. <snip>
However, lessons learned in a workshop like this, would easily carry over to the 970 as well.

ArsTechnica mentioning that the Altivec/Velocity Engine would be the 970s greatest strength tends to argue for optimizing to it.

Once there is a 970 and a G4 - it would be interesting to compare the hardware speed (using Linux on both?).
And then a test of OS X vs Windows XP speeds, to see how good the OSes are.

 

[mathiasr]

Originally posted by GregAussie
Yes ArsTechnica shows that the 970 will not be up to the P4 speed. Except for number crunching/Altivec(Velocity Engine).

Hmmm... I would not draw such a conclusion based on Ars Technica article. They said nothing about the instruction set, these guys dig deep inside the chip and tend to overlook some points. For instance they did not try to compile a simple piece of C code and simulate its execution to actually have an idea in how many IOPs and µ-ops it will break, it's my feeling that the x86 code will generate more of those micro instructions to produce the same work amount.

Many people will tell you that x86 and PowerPC instruction sets are hard to compare. The PowerPC instruction set is based upon the fact that the architecture offers 32 registers of each type (integer, floating point, vector), most instructions have 3 operands (2 sources, 1 destination), PowerPC is also a load/store architecture. The x86 instruction set has only 8 registers (they do not pass their function parameters using registers), and its floating point unit was based on a stack structure, instructions have usually only 2 operands (1 source, 1 source/destination).

You cannot simply ignore these facts when you compare those chips. You cannot compare IPC (Instructions per Cycle) of two different ISAs, without saying that the instructions in question are not the same. What actually does the work is a flow of dozens, hundreds, thousands of instructions... how could you draw meaningfull conclusions when you base your analysis solely on a few single instruction trips through the CPU?
I'm not saying that Ars Technica did not a great job, just that they focus to much on the CPU back-end and do not take the instruction set into account, which would lead to a better overall picture.

 

[the future]

Re: Re: How the hell??

Originally posted by Snowy_River
However, I'm not going to hold me breath...
(Yes, I admit it, I'm addicted to air. Really, I'd try to quit, but the withdrawls are killer... )

Yeah, me too. The Soundtrack to "Virgin Suicides" was really great...

 

[dongmin]

Re: PPC970 at WWDC 2004?

Originally posted by fpnc
As for the various rumors over the last few weeks, I think that many of these have been largely debunked. In fact, the recent Arstechnica paper on the PPC970 should have placed the last nail in the coffin on those rumored performance benchmarks.

...

When the 970's arrive they should be very nice machines but they won't be setting any new price-performance marks in the PC industry. They certainly will represent a new and vastly improved phase for the Macintosh but I won't be expecting miracles (such as dual 970's at less than $2K or single-processor machines that CLEARLY outperform high-end P4-based systems on a wide range of tasks).

Well it's always nice to get a bit of the pessimism to douse the frenzy here. At this point, all the benchmarking and performance estimations are purely speculative. And here I'm referring to both the sensational reports of MacBidouille as well as the more conservative (and technical) opinion of Ars. Yes the MacBidouille benchmarks have a lot of holes but we simply don't know enough about how the 970s will be implemented by Apple to say how it'll perform real-world tasks. We don't know how much the increased bandwidth will affect Altivec performance, which could potentially be huge. We also don't know if Apple is implementing any new bus & interface technologies (RapidIO, Hypertransport, the rumored 'Apple Proccessor Interconnect Bus'). Too many unknowns at this point.

The closest to an 'official' performance estimate we have is IBM's SPEC estimates which show the 1.8 ghz 970 to be roughly on par with the 3 ghz P4 and double the G4's at the same clock speed. This is the main source of optimism for me. I have no doubt we'll see a dual 1.8 ghz system which should surpass the performance of any single CPU x86 system and compete with dual Xeon systems.

On the other hand, such performance levels would be an unprecedented, massive jump over the previous generation.Would Apple make such a bold move? Or would they 'phase in' the performance gain to give themselves some head room for growth and to make people feel better about their G4 purchases? That's my big concern, that Apple would hold back the 970 for non-technical reasons.

And regarding the cost of these systems, I expect Apple to stay with the existing scheme. The 'Fast' system sporting a single 1.4 ghz 970 at around $1500-1600, 'Faster' dual 1.6 ghz system at $2000-2200, and 'Fastest' dual 1.8 ghz system at $2700-3000. Just for kicks, I can see Apple providing a limited run of dual 2 ghz systems with all the bells and whistles, calling it the 'Workstation' config, for $4000. The sole reason for the last system would be for bragging rights--the fastest desktop computers in the planet!

 

[wallinbl]

Re: Re: PPC970 at WWDC 2004?

Or would they 'phase in' the performance gain to give themselves some head room for growth and to make people feel better about their G4 purchases? That's my big concern, that Apple would hold back the 970 for non-technical reasons.
[/B]

Nah. They would just be wasting the opportunity to sell! If the 970 based Mac was 2x as fast as the G4, even more people would ditch their G4 for the 970.

 

[Frobozz]

Re: PPC970 at WWDC 2004?

Originally posted by fpnc
As for the various rumors over the last few weeks, I think that many of these have been largely debunked. In fact, the recent Arstechnica paper on the PPC970 should have placed the last nail in the coffin on those rumored performance benchmarks.

I completely disagree with your analysis. I think the ArsTecnica article just supports our claims for speed. The speed of the machines has a lot to do with the architecture of the final 970 with Altivec, which the author did not have his hands on-- he mainly had Power4 references. Additionally, the speed relies on the architecure of the Computer. He said quite distinctively that, given the right architecture, the 970 will scream.

Apple will be releaseing the 970 based mac no later than the release of Panther... but I really do believe that it could be at, or right after, WWDC. There's just way more logic to support this-- Apple can't afford to sit on their a$$ for much longer.

 

[Snowy_River]

Re: Re: PPC970 at WWDC 2004?

Originally posted by dongmin
...Would Apple make such a bold move? Or would they 'phase in' the performance gain to give themselves some head room for growth and to make people feel better about their G4 purchases? That's my big concern, that Apple would hold back the 970 for non-technical reasons...

There is one point here that I think some people who are afraid of this are missing. What would the public reaction be if the fastest new Power Mac had 1.6GHz 970 processors, but the new IBM Blades had 2.2GHz 970 processors? There would be some monumental negative backlash for Apple if they didn't use the fastest 970s available, and this time Apple isn't the only company that will be using these processors in computers. So, we'll have a touchstone to compare where the machines' speeds are at...