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

"Numerous kernel and library enhancements provide better performance on the same hardware. Smart thread scheduling, in the form of thread affinity and preemption, makes more efficient use of both single and multiple processors."

and also from Apple:

"Jaguar includes enhanced preemptive multitasking, symmetric multiprocessing and multithreading capabilities that speed up individual applications and provide true multitasking capabilities." :cool:

PS- are they 7470's?:confused:
 
Originally posted by Edge100


Of course, the similarly spec'ed PC would blow the doors of the PowerMac, but your point is well taken.

PCs are faster (what that means in the real world depends on the user!!!). But they dont run OS X so they dont get my money.

blow the doors off? hardly. similiarly spec'ed pcs pitted against the old powermacs very narrowly beat out the powermacs (amd(1900+) 20%, intel dead heat against a 2ghz intel, havent seen the benchmark against a 2.5 yet). although id agree a similiarly spec'ed pc would beat these new powermacs i think it would be very very close.

amd yes you can build a nice pc system for less, but you can also configure a barebones powermac for under $1000 if you went to an apple specialist. the reason i have been harping on the similiarly spec'ed thing is that most people here probably have never built a pc before and they beleive these idiots who claim they can build a machine compareable to the powermac for 1200. you cant. you can build a very nice system for 1200 but it still wouldnt have everything the powermac does. if you spec it out part for part and get things as close as possible the machines will cost about the same. the pc quite possibly will beat the powermac in some benchmarks (id bet the powermac woudl also win a few benchmarks) but the difference would be negligable especially when one considers the non-hardware advantages of a mac.
 
Re: Dual processors and Jag

Originally posted by soilchmst
"Numerous kernel and library enhancements provide better performance on the same hardware. Smart thread scheduling, in the form of thread affinity and preemption, makes more efficient use of both single and multiple processors."

and also from Apple:

"Jaguar includes enhanced preemptive multitasking, symmetric multiprocessing and multithreading capabilities that speed up individual applications and provide true multitasking capabilities." :cool:

PS- are they 7470's?:confused:

Apples claims hold as much water as MS's claim that they had near DVD quality at 500kps. (WMV8)
 
Re: Germam prices...

Originally posted by groovebuster
I just checked the prices in the german Apple Store (based on the current exchange course):

DP867: 1845.- US$ + tax
DP1000: 2,620.- US$ + tax
DP1250: 3,490.- US$ + tax
DP1250 (ultimate): 5,470.- US$ + tax

They are really nuckin' futs!!! The rip off goes on for european users! I am really mad now! The DP1250 costs almost 200$ more here! :mad: What do they think??? ... in case they think at all!!!

Yeah, that's how you increase market share Apple!!! Go on like that... good plan!

*shaking head*

groovebuster

If Macs cost more in Europe can't you just buy one from a U.S. website? Or find a friend to purchase one and have it shipped directly to you.
 
Re: Re: Dual processors and Jag

Originally posted by Ibjr


Apples claims hold as much water as MS's claim that they had near DVD quality at 500kps. (WMV8)

then dont buy jaguar. those of us who have used it know it is all apple claims it is, but if you enjoy having your head in the sand then quit bitching and just dont buy it.
 
Originally posted by Ibjr


http://www.xlr8yourmac.com/G4ZONE/QT_imovie_G4tests.html

Wow, i must say! A whole 3 seconds. Not worth the cost.


Those systems were not identical. Example, the single CPU system had a 7200 rpm hdd vs the 5400 in the dual. And, if you've ever done anything video editing wise, you'd realize that it is by far a HDD limited program. Ever wonder why those people take 15k rpm hard drives and run them in raid 0? Hell, even when my friends does editing on his dual athlon it only uses 70% of both cpus because his 10k hdd can't keep up. I would take (and have taken) a slower dual system over a faster single proc system any day of the week.


topicolo: I'm actually a Junior in Computer Engineering at Iowa State.
 
Originally posted by eirik


I don't know. DDR can double the effective throughput. So, even if the bus is at 200 MHz, if your L3 cache runs on an independent bus that runs at 500 MHz, then that L3 cache really provides a significant boost (an extra 100 MHz) in performance for small or many small jobs on your PC, making the most out of your CPU's.

I'm surprised that the x86 boxes don't use L3 cache (I'm assuming you're correct about that.). They have all the more reason than Apple to do so because of their very highly clocked CPU's, which can deliver some pretty impressive throughput. I imagine their CPU's are data starved as well.

I agree, but they didn't remove the L3, just reduced it on the low and mid model from 2MB to 1MB. Think about it. Neither of these are very large, and at 500Mhz they are moving data through pretty fast. If the main memory bus is running faster that might be enough to make up for the smaller L3. My feeling is they do this to make the high-end machine faster than the mid, so it seems to justify the price difference. If the high-end was a dual 1.5 or 2GHz, then they wouldn't have to do that.
 
You could build a PC that would be faster and cheaper than the 1.25. But, you'd be building it with cheap (poor quality) parts which don't always agree with each other.
 
Originally posted by locovaca



Those systems were not identical. Example, the single CPU system had a 7200 rpm hdd vs the 5400 in the dual. And, if you've ever done anything video editing wise, you'd realize that it is by far a HDD limited program. Ever wonder why those people take 15k rpm hard drives and run them in raid 0? Hell, even when my friends does editing on his dual athlon it only uses 70% of both cpus because his 10k hdd can't keep up. I would take (and have taken) a slower dual system over a faster single proc system any day of the week.


topicolo: I'm actually a Junior in Computer Engineering at Iowa State.


just some benchmarks to help with the whole single versus dual discussion.
http://www.barefeats.com/xserve2.html

http://www.barefeats.com/pentium4.html
 
AmbitiousLemon:

"guess you didnt read closely. i always find it interesting how these spec benchmarks show the mac as being 'crushed' by the wintel world and then in real world benchmarks surprisingly you find that the pc isnt actually all that much faster. btw if you cant understand what was written there (its not just marketing speak) then you shouldnt be claiming it means nothing. it might mean nothing to you. pretend as you like, but pipeline length is important. and if you cant get that then i am sorry."

Dunno about you, but I have taken years of university classes in digital+processor design, assembly, OS design, compiler design, programming in general... I work as a programmer and I have worked as a PC and a Mac tech. I also have spent years reading hardware sites, and I read though parts of technical docs from Intel, Moto, and AMD from time to time. So... yes in the end I've got a pretty good handle on this.

Yes, SPEC means more to me than Apple Photoshop "tests" or the speed of a G4 running RC5. The G4 pretty much brings up the rear in SPEC when ranked against other processors that are available in desktops/workstations/servers.

This is not to say that I would pass on an Apple desktop, in fact I'm thinking about getting a 1.25 when they ship. But the fact of the matter is that the G4 is not an especially good general purpose processor (although AltiVec remains very nice).

(Oh, and yes all that stuff you quoted was just Apple's marketing speak, and it does little to impress me.)
 
Originally posted by eirik


I'm surprised that the x86 boxes don't use L3 cache (I'm assuming you're correct about that.). They have all the more reason than Apple to do so because of their very highly clocked CPU's, which can deliver some pretty impressive throughput. I imagine their CPU's are data starved as well.

That's because L3 cache costs a lot of $$$, and is usually found on high end workstation/server chips. A majority of the X86 community is the "cheap boxes" community, and adding L3 cache would make it much more expensive and would provide little benefit for their needs. L3 cache is better suited to databases and other software where you have large datasets you need quick access to. Memory that fast is expensive, especially in the quantities that Apple is using (1-2 megs).

The Intel Xeon MP chip has 512k - 1 meg of L3 cache.
 
Originally posted by cooper13


Sure. Too bad the E3500 with 1 proc starts at $52K and an E4500 with 2 procs starts at $124K!! If we think Apple has bloated pricing, try buying from Sun!

Oh, of course they are. They'll take you to the bank and back again. But their hardware scales and it's balanced. I was just responding to the guy who thought that Macs were the be all end all of performance.

When you get into the markets where Apple looks like they want to target...traditional SGI, Sun, HP, and now Linux markets, their workstations can't compete. People in these markets will pay a premium for extremely fast hardware because...and this sounds so cliche, but...time is money. Either that or they'll get inexpensive (but still up to Mac spec) commodity hardware, run Linux, and have a cost effective render farm. Apple needs to be competitive here and this doesn't look good for them.
 
10.2 I can't wait. Sped up finder, faster overall performance, GFX Card accelerated Aqua... *drool*

Is the finder Cocoa yet or is it still Carbon?

As for my beef with 10.2 is more of a beef with apples business practices O:)

As for the machines, the look nice, and the 867 looks to be the best bang for the buck, even in the long run.

My problem is, that the high end cost so much more. 3299? that's 300 more than the old cost of a dual Gig. They really need to go with the pricing of 1599, 2299 and 2999 for marketting and to keep prices consistant between lines. Those higher prices are going to hurt them more than anything else.

BUT, the fact that all the pro desktops now have duals and the consumers have singles is an excellent dividing line.

Maybe they will go dual in the pro portables? Dual G4's in the Titaniums? :9
 
Originally posted by Ibjr


'Jaguar' should not help SMP, that would be the bsd. (I think, not sure!?!)

OS X doesn't run on BSD, it runs on the Mach kernel. It does have a BSD "subsystem" that gives you all the UNIX tools, but you can actually install OS X without the BSD subsystem as a custom install option.

The kernel is where the SMP happens.
 
ddtlm. :) sorry im not impressed. basically all you said is that despite an education (from where i wonder) that you still will ignore all tests of real world performance i favor of a benchmark that in no way translates to the speed of the processor. furthermore you ignore everything regarding the differences in the archetecture of the chip all to stare blindly at a single benchmark that says something different from every single other benchmark performed on these machines. your education means absolutely nothing to me as long as you keep those blinders up. the plain fact of the matter is that the spec benchmarks do not test real world performance. since this machine is a desktop personal computer real world benchmarks mean far more than spec. what do the real world benchmarks show? well yes the powermacs are slower. but it is a narrow victory for the intel folks (amd does better). these new machines should make the gap even smaller. if you are interesting in using a computer how it is designed to be used apple does well. if you want to run spec all day get a pc and enjoy.
 
I can show proof that my G4/400 is nearly twice as fast as a Dell P4/1.7ghz. Run rc5 on both. Talk about pathetic, how can this poor G4/400 beat the pants off of a P4/1.7?
 
Originally posted by AmbitiousLemon


blow the doors off? hardly. similiarly spec'ed pcs pitted against the old powermacs very narrowly beat out the powermacs (amd(1900+) 20%, intel dead heat against a 2ghz intel, havent seen the benchmark against a 2.5 yet). although id agree a similiarly spec'ed pc would beat these new powermacs i think it would be very very close.

What benchmarks are you talking about here? Steve Jobs' Photoshop bakeoff. Marketing that is. Run a different set of filters that are SSE2 enabled and the P4 will come out ahead.

I'd love to see an independent standard benchmark like SPEC (or TPC-C since I'm a DBA). Granted, SPEC isn't always the best indicator of system performance but it's pretty much the best independent benchmark out there for processor performance. And the G4 doesn't end up in a dead heat against a 2GHz P4 in SPECcpu.
 
Originally posted by Edge100


Of course, the similarly spec'ed PC would blow the doors of the PowerMac, but your point is well taken.

PCs are faster (what that means in the real world depends on the user!!!). But they dont run OS X so they dont get my money.

The ultimate smackdown: Mac versus PC

The ultimate smackdown: Mac versus PC

By Gene Steinberg

The other day someone sent me a message stating that he really liked Macs, but he felt they were too slow. How did he conclude that? Simple. The fastest Mac had a one-gigahertz processor (all right, two of them), and the fastest "Intel Inside" Windows box had already achieved 2.4 GHz.

Doesn't that prove a Windows PC is faster?

That's the impression Apple has struggled to overcome with its "bake-offs," featuring the fastest Mac of the week against a top-of-the-line Pentium 4. The latest comparison I consulted at Apple's Web site claimed the most powerful Power Macintosh G4 was 68% faster than a 2.2-GHz Pentium 4 in a series of test runs with Adobe Photoshop "using nine commonly used actions and filters that stress overall system performance."

Testing the claim
Rather than simply select a PC hotrod for the ultimate smackdown, I looked for a Windows box with a touch of class, since Apple is so heavily involved in style. I came up with a Sony Vaio RX690G Digital Studio, which incorporated the 2.2-GHz processor. Now I should point out that things move awfully fast in the PC universe, and this model has since been replaced with a RX790G, where processor speed is boosted to 2.4 GHz, but I'll await the next Power Macintosh upgrade to do that comparison.

Putting the pedal to the metal

Apple's test protocol suggests the test computers run with as few system encumbrances as possible, and that includes disconnecting network cables. But because graphic artists don't usually work in that fashion, I decided to keep the configurations normal, even if it hobbled potential performance somewhat. The Mac was upgraded to the latest version of Mac OS X, 10.1.5. The Sony had Windows XP. I installed the standard retail versions of Adobe Photoshop 7 on both computers. Apple's original comparison employed an older version of Photoshop, which no doubt accounts for some of the differences between my test results and theirs.

Running the tests proved exceedingly simple because Photoshop displays the actual timing of a rendering process rounded off to tenths of a second. Per Apple's directions, I conducted each test four times to deliver the most accurate results.

Like all Adobe applications, Photoshop is a bit slow to launch. It took 15 seconds on both computers to get ready for the main event.

In the nine test runs, the Mac came out on top five times, besting the Sony by up to 8.1 seconds. Where the PC emerged victorious, the margin was usually less than half a second.

In all, the Mac took a total of 35.5 seconds to complete the nine rendering steps. The PC took 50.1 seconds, making it 41% slower according to my calculator's reckoning.

All right, as that old song says, I'm a believer.
 
Here is a similarly specced DIY Athlon MP system :

2x$122-- Athlon MP 1600
$167-- Tyan S2460 Dual Athlon MP/ATX mboard
$111-- PC2700 DDR 512MB
$149-- ATI Radeon 9000 pro
$128-- Sound Blaster Live Platinum 5.1
$60-- Belkin 1394 firewire controller
$50-- D Link Gb Ethernet
$250-- Pioneer DVD-R drive
$113-- case
$20-- keyboard
$20-- mouse
$299-- Windows XP Pro
$84-- EIDE 80.0GB
_________________
$1695

There are some other parts that are required to build this system from nothing so the price is approximate. prices pulled off of pricewatch


The Prices on the new PowerMacs aren't that outragious. After all, they are workstations, not your run-of-the-mill consumer PC. And I'd like you to price of a brand name Athlon MP or Intel Xeon workstation. The price will be very close.

Like another poster stated earlier, if you think the systems are too expensive, your not Apple's target market, and you don't have to buy one.

And those of you who are saying "I'm waiting till MWSF" I wouldn't get your hopes up. With a release this late in the year I'd be surprised to see a blow your socks off update until at least MWNY.
 
Originally posted by kenohki


What benchmarks are you talking about here? Steve Jobs' Photoshop bakeoff. Marketing that is. Run a different set of filters that are SSE2 enabled and the P4 will come out ahead.

I'd love to see an independent standard benchmark like SPEC (or TPC-C since I'm a DBA). Granted, SPEC isn't always the best indicator of system performance but it's pretty much the best independent benchmark out there for processor performance. And the G4 doesn't end up in a dead heat against a 2GHz P4 in SPECcpu.

no im not talking apple's benchmarks if i was id say the powermac is 90% faster than a 2.5intel. benchmarks are available all over the web from many independant sources. i quoted some above. every source i have found indicates someting similiar. the old powermacs were down about 20% from the top of the line amd. and about even with a 2ghz intel (i havent seen benchmarks against a faster intel).
 
Originally posted by rugby
I can show proof that my G4/400 is nearly twice as fast as a Dell P4/1.7ghz. Run rc5 on both. Talk about pathetic, how can this poor G4/400 beat the pants off of a P4/1.7?

Because RC5 on the G4 is using the SIMD unit (read, AltiVec) which allows it to process lots of data in parallel on that specific task.
 
Originally posted by AmbitiousLemon
ddtlm. :) sorry im not impressed. basically all you said is that despite an education (from where i wonder) that you still will ignore all tests of real world performance i favor of a benchmark that in no way translates to the speed of the processor. furthermore you ignore everything regarding the differences in the archetecture of the chip all to stare blindly at a single benchmark that says something different from every single other benchmark performed on these machines. your education means absolutely nothing to me as long as you keep those blinders up. the plain fact of the matter is that the spec benchmarks do not test real world performance. since this machine is a desktop personal computer real world benchmarks mean far more than spec. what do the real world benchmarks show? well yes the powermacs are slower. but it is a narrow victory for the intel folks (amd does better). these new machines should make the gap even smaller. if you are interesting in using a computer how it is designed to be used apple does well. if you want to run spec all day get a pc and enjoy.

The SPEC benchmark tests one thing: CPU speed (and to a lesser degree, memory bandwidth/effiency). In many computing applications, this is not the bottleneck, but rather HDD speed, OS effiency and memory management, and the chipsets, to give a few examples. While an overkill, you could take a 7200 rpm hdd and put it in the original Macintosh and it would fly- up to a point, but do you catch my drift?
 
u yes i 'catch your drift' but what you just said is simply proving my point? if you are trying to agree with me um thanks but no need to be snotty with the catch my drift bit. i appreciate someone else trying to educate all the troll we have today.

as long as you are proving my point you should also point out that while spec tests 'cpu speed' it does so only under imaginary conditions. its sort of like when intel claims usb2 is faster than firewire. ist faster on paper but you drop it into the real world and usb2 seriously lags behind fw. you take a machine that cranks through a spec benchmark with high ratings and drop it into the real world and you will see it choke. sure it might still beat a powermac but the powermac will be in the race. none of this mac gets blown away nonsense some people like to claim.

loc. :) np. i wasnt sure if yo were agreeing with me or not.
 
Originally posted by AmbitiousLemon
u yes i 'catch your drift' but what you just said is simply proving my point? if you are trying to agree with me um thanks but no need to be snotty with the catch my drift bit. i appreciate someone else trying to educate all the troll we have today.

Sorry, I wasn't trying to be snotty, I was just trying to reinforce your points. Thought the best way would be to add to what you said, guess it came off wrong.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.