check this article out regarding recent benchmarks...
http://www.theregister.co.uk/content/3/24358.html
http://www.theregister.co.uk/content/3/24358.html
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.
This looks bad for apple
- Thread starter OSUbuckeyefan
- Start date
- Sort by reaction score
As bad as it looks for Apple, there is more than one side to it. Look back in February for a previous thread with a very good discussion of the flaws in the testing. This is not "new" news.
Edit: here's the first time around. (mrtrumbe: my sentiments exactly!)
https://forums.macrumors.com/threads/2413/
Edit: here's the first time around. (mrtrumbe: my sentiments exactly!)
https://forums.macrumors.com/threads/2413/
I know exactly what you mean.
No fanatical rantings here about 400MHz G4s running rings around 2.2GHz P4s.
OS X still, apparently, has some inefficiencies in the code. I'll tell ya, Linux is very fast. Why is that Unix so much faster than what Apple's using? My B&W 350MHz G3 runs OS X, but noticeably more slowly than OS 9. OS 9 feels looser than an oily bar of soap now in comparison.
I believe the other thing the benchmarks revealed was that Moto's compilers were lacking as well. So it's not just the OS.
No fanatical rantings here about 400MHz G4s running rings around 2.2GHz P4s.
OS X still, apparently, has some inefficiencies in the code. I'll tell ya, Linux is very fast. Why is that Unix so much faster than what Apple's using? My B&W 350MHz G3 runs OS X, but noticeably more slowly than OS 9. OS 9 feels looser than an oily bar of soap now in comparison.
I believe the other thing the benchmarks revealed was that Moto's compilers were lacking as well. So it's not just the OS.
This has popped up in 4 threads.
It's trash. The test is poorly designed and has no isolation for other hardware differences like mobo speed and RAM. I'm getting real sick of seeing the same flawed study.
It's trash. The test is poorly designed and has no isolation for other hardware differences like mobo speed and RAM. I'm getting real sick of seeing the same flawed study.
So tell me, Mischief...
...how would you conduct such a test to be fair? I know nothing about all this benchmarking stuff.
...how would you conduct such a test to be fair? I know nothing about all this benchmarking stuff.
Not a lot of comments THIS time, because lots of comments have already been made several other times this thread started. Thus the reason for mrtrumbe's expression of dismay. Look back at my previous post for one of the earlier threads. This has been talked to death on this and other boards.
Grrrr!!
The reason I said, "Not again!" is that this very issue (and this very article) have appeared on these very forums before. I personally have commented extensively on what I think about TheRegister's comments.
Basically, their comments on OS X are laughable simply because OS X's slowness is completely unrelated to the benchmarks. Also, there is some merit in saying floating point (double precision mind you) operations are not as fast as they should be on the G4. However their fp tests in general are biased for reasons including, but not limited to, their lack of utilization of the Altivec unit.
For all of my comments on this article, go here: https://forums.macrumors.com/threads/2845/
BTW- This is about the third or fourth time the C't benchmarks have been brought up on these forums. Don't complain about the lack of comments. Everythings already been said.
Matthew
The reason I said, "Not again!" is that this very issue (and this very article) have appeared on these very forums before. I personally have commented extensively on what I think about TheRegister's comments.
Basically, their comments on OS X are laughable simply because OS X's slowness is completely unrelated to the benchmarks. Also, there is some merit in saying floating point (double precision mind you) operations are not as fast as they should be on the G4. However their fp tests in general are biased for reasons including, but not limited to, their lack of utilization of the Altivec unit.
For all of my comments on this article, go here: https://forums.macrumors.com/threads/2845/
BTW- This is about the third or fourth time the C't benchmarks have been brought up on these forums. Don't complain about the lack of comments. Everythings already been said.
Matthew