Preliminary OS X & PPC 970 Benchmarks

Dixie_Flatline writes "Macbidouille.com is reporting that they have preliminary benchmarks involving PPC970 hardware. The results are seriously impressive. We're looking at a single processor PPC 970 1.4GHz machine quite strikingly outperforming a dual G4 1.42GHz machine. Don't worry, there's an English translation embedded in the page so you don't have to try to muddle through the French." Update: 05/05 19:58 GMT by T : Thanks to Eric from macbidouille.com, above link updated to a static page; hopefully you'll get better response this way.

Comments

They almost got it. . .

[noewun]

Inneresting. However, I wish they would have left off the

Mac fans, our wait will be rewarded. The fight is over and Apple will soon rule the world !

cause it makes the whole article sound silly. I've been a Mac user since 1989, but I really, really, really, really wish people would find something more interesting to argue over than which platform/OS you use.

Nice, but still skeptical....

[Dr. Mojura]

While I truly believe Apple will use the 970, and I'm sure it will be much faster that their current offerings, I still have to remain skeptical of this. Call me naive, but how am I to believe they not only have alpha releases of panther (very possible, since they are probably developer seeds), but they also somehow obtained unreleased hardware as well? "...were done on the same computers that will be sold." I can't imagine Apple is so loose to let out alpha/beta harware.

Then again, never underestimate the marketing power of 'hype'. Whether it's true or not, I hope the release is sooner than later.... I miss OS X.

Re:So it is faster than dual G4s

[dAzED1]

did you read the article at all?

Right there in the article is a comparison to the P4 3Gz, which the 970 is only slightly faster than. Now, compare the price of a p4 3.0 (or a duel p4 2.6 or such) to the price of a 970...

Glad that there is a great chip out there (970) but price per performance, the guy (intel) that's making far more chips still is doign it cheaper. Economics.

Remember the "Apple" section of /.?

[chasingporsches]

Maybe this should have been posted to the Apple section. I had to actually click on the /. logo to see this post! My home page is the apple section, to filter only what i really want to see ;-)

Re:Apple Secrecy Sucks

[Valdrax]

Perhaps you should read the educational tale of the Osbourne to learn exactly how your reaction is exactly why Apple keeps this kind of thing secret:

Two years after the Osborne débuted in the marketplace, the company was bankrupt! Why? Osborne made a fatal mistake! They failed to plan! To make matters worse, they announced a second machine, the Osborne 2, which was suppose to be a great improvement over the initial Osborne and all of its competition. However, the announcement was premature since the Osborne 2 was not ready for customer shipments. As a result, the sales for the original Osborne dried up while their customers waited for the Osborne 2. Consequently, the company had no sales that translated to no revenue and subsequently to no cash. Meanwhile, the market was waiting for the new and improved Osborne 2 that never materialized because the company ran out of money, and went bankrupt. The company failed in large part to a lack of honest and intensive business planning. In summary, by failing to plan, it plan to fail and was found guilty of "eating its own children."

-- From this site.

Re:Mac Rumour Mongering

[John Harrison]

Also remember that Steve likes to announce things himself. There is not way he would stand for a leak of this size. The chances of this being real are slim indeed. If it were real they would show us a picture of the machine and its innards. That said, if it is real this might be enough to get me to purchase another Mac.

Re:Sad...

[Steveftoth]

They're not running in 64-bit mode.

The 970 runs 32-bit PPC code and 64-bit PPC code. Only the kernel has to support the 64-bit mode and switching back and forth for the individual applications.

However, these benchmarks still might be fake. It's hard to tell since I can't even download the article.

Re:Mac Rumour Mongering

[Josuah]

Remember that MacBidouille has a history of inaccurate rumors... remember their AMD rumor earlier this year. Check out their rating at www.macrumors.com

I find it very hard to believe that a rumor consists of hardware benchmarks.

Pot calling the kettle black

[SuperBanana]

Remember that MacBidouille has a history of inaccurate rumors... remember their AMD rumor earlier this year. Check out their rating at www.macrumors.com

Yeah, cause, you know, those MacRumors guys are real grounded. Fact is, almost the entire crowd of Macintosh rumormongers NEVER get it right. There's usually the slimmest glimmer of commonality between what they claim, and what actually happens.

Remember the "definative" pictures of a new "Apple PDA", supposedly sitting on someone's desk at Apple, that turned out to be a complete hoax? Apple buying up a music company? The list goes on. These guys take a sniff of one little piece of info(like, maybe Apple execs meeting with music industry execs) and spin it into the most preposterous fiction(Apple buying a dying music company, supposedly for its wares). Instead of looking at their track record, everyone just keeps paying attention to them...which is stupid, because their dreaming is always more grand than the rabbit Steve pulls out of the top hat.

Fact is, the AMD rumors were the result of a numbskull who decided to get some free publicity through lying-by-omission-of-detail(summary: "Are you talking to Apple?" "We'll have to get back to you on whether we are allowed to talk about how we're talking to Apple.")

Re:Pot calling the kettle black

[Brandon Sharitt]

Well MacRumors doesn't really release its own stories that often. Most of the time they just post links to sites like MacBidouille and tell you to take it with a grain(or barrel) of salt.

Re:So it is faster than dual G4s

[Shuh]

did you read the article at all?

Right there in the article is a comparison to the P4 3Gz, which the 970 is only slightly faster than. Now, compare the price of a p4 3.0 (or a duel p4 2.6 or such) to the price of a 970...

Now consider the 1.4GHz single-processor 970 is goint to be the bottom-of-the-line PPC tower.

Don't trust

[humina]

I don't trust this information at all. There are a lot of apple people that would like to see a processor that is significantly faster than intel's offering (I am one). Unfortunately there are people that will publish rumors that apple is going to do this soon without proof because they wish it were so now. The only apple rumor site that I would trust is Think Secret. Other than think secret or an announcement from apple, I refuse to believe that any of this information is true. This is merely wishful thinking.

Re:So it is faster than dual G4s

[timeOday]

Well, if you'd looked at the bar charts in the artcle, you'd have seen that the 1.4 GHz benchmarks at about the same or a little faster than a 3 GHz P4.

The 1.4 PPC 970, that is, not the G4 1.4.

Of they G4, they write,

By reading these benchmarks you'll understand that we couldn't publish them before. Now we know that PM G4 sells are stuck at a very low level, the following test results won't have much incidence. It will however make the ones switching to PC wait for the next generation of Power Macs.

Now, maybe I'm reading too much into a rough translation, but it sounds like they were witholding benchmarks that showed how the single P4 3.0 spanked the dual 1.4 G4. That doesn't seem very forthright.

Meanwhile, comparing *today's* Intel product against *tomorrow's* PPC must also be done with caution; by the time you can buy that PPC 970, Intel and AMD will have something else, too.

Macintosh Nerd Factor @ All-Time High

[Shuh]

First they moved to a modern RISC-based ISA in 1994. Then they moved to a UNIX/NeXT-base OS with OSX in 2000. Now they're moving into a Power-4-workstation -derived 64-bit processor that will come out of the gate (at its lowest clockrate) neck-n-neck with the highest clockrate x86 CPU's in their prime.

Throw in things like brilliant X11 support, a desktop graphics subsystem only dreamed about for other OS's now, and even a Nightly Phoenix/Firebird build for OSX.

It's going to be a great time to upgrade a Mac, or buy one if you don't alreay have one.

Re:Macintosh Nerd Factor @ All-Time High

[Anonymous Coward]

Bzzt nice try.

This is being typed on a G4 Cube, what many Mac-bashers have called one of Apple's biggest folleys in years.

And Jaguar runs very well with 576 meg RAM and a first-generation G4-450. Never skips in iTunes. Menus are instantly responsive, system menus work when apps hang, etc.

The worst thing that happened was a UI-freeze after my 802.11 base station froze up.

Loading OS9, OTOH, I find that suddenly, menus take forever to pop up (yay for non-premptible multitasking), my MP3's skip when the system goes for I/O.. It's idiotic.

What you're probably seeing is a bunch of people trying to run Classic. Let OS9 die, for the love of God. Let it die. Don't run Classic apps, they bog the system down. Just let it die.

If you don't have Carbon or Cocoa apps that suit your needs, don't even bother booting into OSX. Stay in OS9 and get some new software contracts. Then LET IT DIE.

Yes, OS9 is faster. I get better performance playing Diablo 2 in OS9, but only a few fps. That's the price you pay for a modern, well-behaved operating system. Deal with it. It's still fast. More than fast enough.

Re:Macintosh Nerd Factor @ All-Time High

[nudicle]

I'm a long-time mac user (1986) and use my 10.2.5 TiBook every day for all kinds of stuff and I have to say that for all the coolness in Apple's recent steps, I think they lost something important in the coherent, simple UI design arena. This is most evident in applications like Addressbook and iCal. Addressbook should win an award for the crappiness of its UI design and iCal suffers from these weird idiosyncracies that in the aggregate really affect productivity. For instance, when adding a new event in iCal, a multi-line text field appears in the window used to create and describe the event. So for weeks I would add events and thoughtlessly press when I wanted to go to the next line ... no problem, right? Well, actually, Apple UI people decided that in this multi-line text field was actually going to perform a .. which means instead of moving to the next line in the field, the entire line is selected and deleted when I, operating on 16 years of mac UI instincts, continue typing under the assumption that my mac behaves like it should. Although I'm kind of used to it now, this was an indescribably irritating quirk to have to get used to and one that I don't think would have ever been deemed acceptable in years gone by. Sometimes I wonder if Apple hired a bunch of UNIX guys to make OS X happen and an unfortunate consequence of that is we have to put up with UI inconsistencies that bleed in from other worlds. I mean, Addressbook introduces the UI "feature" of selected text that is not deletable by the "delete" key (eg when creating a new entry and you're presented with "First" (as in name) .. and the text is selected ... and you can't delete it with "delete." I guess the justification is that the field only accepts legally displayable characters for a first name but I don't care. The fact is that behavior is wrong and it breaks the harmony of using macos.

Don't get me wrong, I love macos x, but I think if Apple knows what's best for it someone will take the GUI people outside and slap them around for a while until they understand that one of the reasons using classic macos was pleasant for so many people is because there was an underlying harmony in how stuff worked

What about a dual PIV?

[The Analog Kid]

Why didn't they test a dual PIV, why cause it might just win? They only tested a single PIV, this bench is just FUD for the Mac. Its uneven and unbalenced, when you test a dual PIV then I'll consider it legit . The gains are only marginal when comparing single processors. I would love to see a canterwood go up against this.

The benchmarks are bogus

[Hannibal_Ars]

We've had some discussion of these in the Ars Mac forum, and the consensus is that they're bogus. I'm currently wrapping up part II of my 970 article, and I'm pretty certain that these numbers are made up.

Here's how it will break down clock-for-clock:

Floating-point: the 970 will spank the G4e
Integer: The G4e will spank the 970
Vector: it's a tie, even though the 970's Altivec hardware is inferior to that of the G4e. What gives the 970 a boost is Dual-channel DDR400 and a real FSB. If you were to put the G4e in a similar system, it would out perform the 970 clock-for-clock pretty handily.

Anyway, I could elaborate more, but I'd rather work on my article.

Re:The benchmarks are bogus

[Hannibal_Ars]

What benchmarks? There are no publicly available benchmarks for the 970 that show anything, aside from whatever numbers IBM's PR dept. released as "projected" SPEC numbers and whatever numbers some Mac rumor site cooks up to drive up traffic.

Re:I wouldn't be at all suprised...

[perky]

Catch an IBM'er and have a frank discussion sometime. And you'll find that the prevailing attitude towards microsoft there is: "One day, maybe not soon, but one day... we WILL bend gates and his minions over a barrel and assrape them HARD. And as they say: 'Revenge is a dish best served cold'".

IBM employs, what, a quarter of a million people? I think you will find that most of them don't give a shit about "assraping" microsoft as you charmingly put it. You are talking a load of bollocks and you are deluded if you think IBM are going to somehow speed up development of a processor, "just to spite Gates". IBM, like every other company, do things to make money and not to spite people. IBMers, like employees of every other large company, care about their project, and maybe even care about the company as a whole; but they don't have vendettas on behalf of the company.

Re:IBM and AMD

[pmz]

Funny, the 970s MHz of 1.4 and 1.8 are equal to the MHz of the new Opterons. Also, isn't IBM building the new Opterons for AMD?

This probably has more to do with the manufacturing processes used rather than some weird conspiracy theory about the Opteron and the 970.

Regardless, the Opteron and this new PPC chip are a damn good thing for 64-bit computing. The Opteron appears to hit a real sweet spot for price, performance, and reliability featurs--let's hope the 970 will do the same.

Re:So it is faster than dual G4s

[gig]

I think they mean that these benchmarks are really for PC users. If you are about to buy a new P4 and do MS Windows all over again then they are saying to you "wait a tick and see how the G5 looks first". There is a widespread feeling in the Mac community that the G5 will be something special because it will be from top-to-bottom all New Apple (post NeXT purchase). There won't be anything at all left from the pre-1997 Mac, basically, that hasn't been completely rebuilt (OS, form factors) or replaced (ADB with USB, A/V with FireWire, CRT with LCD). Also, there is a feeling that IBM made the PowerPC 970 just for Steve and Avi and John, so whatever they wrap around it will take really good advantage of it, and basically blow the doors off Wintel for all you guys who are still stuck with 1990's-era computer systems. The idea is that if you bristled reading the previous sentence they hope to show you that it is really true, with 64-bits and system design that goes from the tiniest hardware element all the way to single pixels on the 3D alpha-composited display.

If you are a Mac user you probably don't care about bar graph benchmarks. I am one and I don't. I just buy a new Mac every three years when my AppleCare is up and sell the old one for half what I paid for it originally and just laugh and laugh.

Mac users are more interested in feature lists like Rendezvous (zero conf networking), FireWire (hook up lots of disks and cameras real fast and easy), CoreAudio (flexibly utilize pro audio interfaces, applications, effects, and instruments simultaneously in real-time), CoreMIDI (route MIDI performance data between applications in real-time), SuperDrive (read and write CD and DVD), Unicode typography throughout, one-crash-per-year stability, etc. And of course I want it in an enclosure that is 50% of the volume of the last one, too.

Bar graphs of a particular render or a particular step or action are fairly useless in creative work. You get a better idea by just using the machine you plan to purchase for 20 minutes at an Apple Store or similar dealer. As long as it has all the necessary features (some noted above) and it feels good then you are set. Apple's Photoshop shoot-outs are not so bad because they run numerous day-long scripts on Photoshop on both platforms. These are scripts that it literally takes an artist all day to record in Photoshop, and you can play them back as fast as the machine can manage, so if you play back 20 scripts on both machines and one is consistently faster then that might be interesting. Not enough to make me ignore how much I don't want to run Windows, though. It's not worth if for so many reasons, not the least of which is the dearth of good creative software on MS Windows.

Re:The 970 vs The Opteron x44 ....... tsarkon repo

[Aqua OS X]

First off, most mac users have known that their platform has been behind the curve (in processing power) for a few years now. I don't know why you'd have to "try to tell" Mac zealots this. Mac users use Mac's because of the better usability, better UI, better hardware integration, etc. I could care less if my new Mac is 8months slower then a new Windows/Linux box.

Secondly, the 970 is far from vapor. It was first presented 6 months ago and they are now rumored to be falling off Hon Hai assembly lines. Not only was the PPC970 announced well -after- AMD and Intel's consumer 64bit solutions, it will most likely be the first 64bit CPU to appear in consumer desktops and laptops.

And finally, what good would an Opteron be to Mac users? Although Cocoa apps could probably be recompiled for a different CPU with minimal headaches, Carbon apps do not port well. Apple would have to create an emulation layer for Carbon apps. It would be a nightmare, it would take for ever to develop, there would be countless software incompatibilities at first, and Mac developers would throw a hissy fit. Shess, we're still coming out of a -major- OS migration.

I could go on and on about why an Apple AMD box would be technically impossible at this point in time...but hey, just trust me, ain't gon'a happen. The PPC970 is a smart move.

Re:The 970 vs The Opteron x44 ....... tsarkon repo

[Ponty]

I'm going out on a limb, but I think you might have anger issues.

Calm down, man. It's a company with products and a CEO.

Not serious

[Bulln-Bulln]

This ''benchmark'' is really bullsh*t.
When you do serious benchmarks, you post details about the hardware and the used OS.
Well, they gave a few details about the Mac they claim to have. But what about the P4-PC?
What kind of RAM did they use? 100MHz? 800MHz? Something in between?
Which Windows version did they use? Was Hyper Threading enabled?
The list could be a lot longer, but you get the point.

Also: Wasn't the PPC970 meant to be a competitor to Intel Xeon and AMD Opteron CPUs rather that just the plain P4 (by price and aimed market)? (I'm not sure about this one.)
Why didn't they benchmark these as well? (They could at least get a Xeon, an Opteron is harder to get.)
The last sentence (''The fight is over and Apple will soon rule the world!'') gives me an indication why they didn't do this: They seem not to be interested in an objective comparison.

Re:PPC 970 on laptoshes

[dadragon]

I'm guessing they're working on a 970 based 15" PowerBook now. The reason I believe this is that the 15" is a good laptop, but there hasn't been a new one in a while, so they're probably keeping it under wraps until the 970 is announced, with a desktop and an Aluminium 15" to go with it.

Re:Hold on a minute

[Tumbleweed]

1) We have no idea if this is even remotely real. I sure _hope_ it is, though.

2) Apple doesn't make PPC970 chips - IBM does. Apple's markup on the hardware is enormous - that's where they make their money. PPC970 machines, if they take over for the G4, will probably be around the same price point. Apple can easily absorb any extra cost of the PPC970 (keep in mind that as die size shrinks, so too does cost - so new generation processors get smaller as well as cheaper to make).

3) The same people that can afford $5k Apples in the past will gladly do so again to get these machines, and those of us who could have afforded them in the past but didn't bother due to slow-ass hardware will also jump ship. And when the PPC970 makes it into the lower-end Macs which will cost >$2k, then lots of others will, too. Why you think you have any insight into how Apple prices their hardware is quite beyond me.

4) If I was gonna buy an x86 machine, it'd be an Opteron to run Linux on, not a Xeon, and besides, Linux doesn't compare to OS X - vastly different user experiences. VAST.

Re:So it is faster than dual G4s

[afantee]

>> But I switched to Linux on Intel five or six years ago and haven't looked back since. As a programmer, I love the tools. As a user, I love the independance. And as a man who appreciates freedom, I savour the chaos, the energy, and the source code.

With OS X, the Mac world is so much better and more exciting than it was 2 or 3 years ago, so you experience gained 5 or 6 years ago is simply irrelevant.

As an long time UNIX and Windows programmer, I can tell you that OS X is truly a dream platform - better than Solaris, Windows and the old Mac OS in every way. I am now much more productive on my $999 iBook than on a $15K Sun Ultra Sparc machine.

Many UNIX and Linux geeks have switched to Mac OS X, including people like James Gosling, Bud Tribble, James Duncan Davidson, Tim O'Reilly, and most of the Perl 6 core team. At least 4 or 5 Slashdot editors are now Mac users.