Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
PCMag's PowerMac G5 2.0GHz Review
- Thread starter MacBytes
- Start date
- Sort by reaction score
You just spent $2K for a computer. Stop bitching and spend another $30 for a mouse you like and $30 on a keyboard you like.
The two-button mouse that comes with a PC sucks too. I throw that out and get a 5-button or 10-button Logitech/Kensington mouse. The huge keyboard that sits in front of my Gateway sucks too, with large buttons like "Email" "Web", that are retarded.
The two-button mouse that comes with a PC sucks too. I throw that out and get a 5-button or 10-button Logitech/Kensington mouse. The huge keyboard that sits in front of my Gateway sucks too, with large buttons like "Email" "Web", that are retarded.
Apple keyboard/mouse are lacking
Is there a decent non-Apple alternative? The keyboard on my iBook is okay, tho' I prefer the Happy Hacking keyboard on my Suns.
For the basic things my wife does with a computer I'm sure she'll be happy with a unibutton mouse when she gets a Mac (which I mentioned to her just before reading your post). It make me uncomfortable just watching her fumble with a two-button mouse on Windows.
Maybe it's time Jonathon Ive started designing Apple keyboards? Or if he already does maybe it's his worst best-kept secret?
Panther improvements?
Where do folks expect to see speedups with Panther?
I can't imagine that Adobe is relying on the Apple frameworks to do the heavy lifting of Photoshop... I'm sure those algorithms are all custom coded.
No industrial app uses the standard OS memory alloc routines, they'll just grab a big chunk once and dole it out in a way that's efficient for the particular app, at least for Carbon apps.
I guess it'll make more memory available, but benchmarks like these aren't using it.
I can imagine Panther speeding up the GUI and system utility applications, but I don't know how it'll change these kinds of benchmarks appreciably...
Improvements will have to come by optimizing the individual apps, right?
Where do folks expect to see speedups with Panther?
I can't imagine that Adobe is relying on the Apple frameworks to do the heavy lifting of Photoshop... I'm sure those algorithms are all custom coded.
No industrial app uses the standard OS memory alloc routines, they'll just grab a big chunk once and dole it out in a way that's efficient for the particular app, at least for Carbon apps.
I guess it'll make more memory available, but benchmarks like these aren't using it.
I can imagine Panther speeding up the GUI and system utility applications, but I don't know how it'll change these kinds of benchmarks appreciably...
Improvements will have to come by optimizing the individual apps, right?
Yeah, I was a little disturbed by the comment that when using the smaller image files both platforms completed too quickly to time...
Were they timing with a stopwatch?!
Very sloppy...
Re: Re: Re: wha wha?
You act as if each instruction has to wait until the previous instruction has been retired. Haven't you realized that the whole point of pipelining is to ensure that instructions don't need to wait to be fetched/decoded/executed/retired before the previous instruction is retired?
Really, the only time a Pentium 4 is going to be at a disadvantage to a shorter pipelined cpu (i.e G4) is when you have dependent instructions and even than the difference is rendered meaningless by the Pentium 4's much larger 126 instruction Out of Order Window (compared to the G4's 16 instruction OOW).
You act as if each instruction has to wait until the previous instruction has been retired. Haven't you realized that the whole point of pipelining is to ensure that instructions don't need to wait to be fetched/decoded/executed/retired before the previous instruction is retired?
Really, the only time a Pentium 4 is going to be at a disadvantage to a shorter pipelined cpu (i.e G4) is when you have dependent instructions and even than the difference is rendered meaningless by the Pentium 4's much larger 126 instruction Out of Order Window (compared to the G4's 16 instruction OOW).
Re: I´ll have 1100 of those...
In the real world any waiting sucks while you're trying to work right? Waiting for tools to load is still waiting.
If they had the G5's energy saver thingy on auto it would be slower.
In the real world people would load the machines up with Ram, they would be multitasking, they would be working with larger images.
I'm glad to see that the pc mag gave any kind of a nod to the G5 and no test can be run where someone won't cry foul, but it is clear they were not interested in really pushing these machines to see what they would do in the real world.
Just the tip of the iceburg for the G5!
daniel
In the real world any waiting sucks while you're trying to work right? Waiting for tools to load is still waiting.
If they had the G5's energy saver thingy on auto it would be slower.
In the real world people would load the machines up with Ram, they would be multitasking, they would be working with larger images.
I'm glad to see that the pc mag gave any kind of a nod to the G5 and no test can be run where someone won't cry foul, but it is clear they were not interested in really pushing these machines to see what they would do in the real world.
Just the tip of the iceburg for the G5!
daniel
Correct me if I´m wrong but doesn´t the performance table clerly state that the Dell is faster overall in the PS7 tests?
PCMag just timed the individual filters and and added up the scores, which is a very flawed method of testing. You should normalize the score. Let's say a single filter is slower on the Mac than the PC, but it's a filter that takes a long time to complete. If you just add up the total time, that one filter affects the score by an uneven amount. So you normalize the score, which is what the folks at Ars Technica have done.
Also, PCMag didn't include three filters that are much, much faster on the Mac - Despeckle, Pointillize, and Radial Blur, which are all included in the PSBench7.
PCMag just timed the individual filters and and added up the scores, which is a very flawed method of testing. You should normalize the score. Let's say a single filter is slower on the Mac than the PC, but it's a filter that takes a long time to complete. If you just add up the total time, that one filter affects the score by an uneven amount. So you normalize the score, which is what the folks at Ars Technica have done.
Also, PCMag didn't include three filters that are much, much faster on the Mac - Despeckle, Pointillize, and Radial Blur, which are all included in the PSBench7.
Re: Re: Re: Re: wha wha?
Incorrect. Xeon, like Celeron, is a bit of a "whatever fits" nameplate, and have in the past been based off the P3 architecture. However, current Xeons are based on the current P4 design (with the multi-proc bits added in, of course).
Incorrect. Xeon, like Celeron, is a bit of a "whatever fits" nameplate, and have in the past been based off the P3 architecture. However, current Xeons are based on the current P4 design (with the multi-proc bits added in, of course).
well this is from a guy who is all PC
wife is on a Mac
I started on Macs from the first 128K one and was always smiling when I used it and went through many Macs after that but around the win95 area I switched as I felt Mac lost it
I want that smile back when I use a puter
anyway I am a photographer for a living and the one thing funny was the guy said that if they included the load times the Mac would have won ??? well idiot put the load times in there should have put in start time and end time etc....
I am truly happy the Mac will now have a fast machine again and hope to be getting one here in the not to near future
but I will still work on a PC for some aps such as C1
also true about the Xeon is a p4 core also the new p4 3.2 extreme is a Xeon
anyway fun to read the tests for sure
wife is on a Mac
I started on Macs from the first 128K one and was always smiling when I used it
and went through many Macs after that but around the win95 area I switched as I felt Mac lost it I want that smile back when I use a puter
anyway I am a photographer for a living and the one thing funny was the guy said that if they included the load times the Mac would have won ??? well idiot put the load times in there should have put in start time and end time etc....
I am truly happy the Mac will now have a fast machine again and hope to be getting one here in the not to near future
but I will still work on a PC for some aps such as C1
also true about the Xeon is a p4 core also the new p4 3.2 extreme is a Xeon
anyway fun to read the tests for sure