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

jEdit performance on Intel iMacs?

D

dolphin842

macrumors 65816
Original poster
To anyone who has a new Intel iMac and has used jEdit in the past... how well does it work under Rosetta? The UI can be a bit sluggish on a PowerBook so I'm wondering if it's unusable or not.
 
jEdit wouldn't need to run under Rosetta, since the Apple JVM is native for Intel (i.e. they converted it to intel). So I would venture to guess that jEdit will be faster...

I wonder if anyone has done any java based benchmarking comparing the JVMs??? Would be sweet if Eclipse was a bit more responsive - it can be a dog sometimes.
 
Cool... that's true Jobs did say that Java apps wouldn't have a problem... didn't even think of that!

I'd be interested as well in seeing how java performance compares to PowerPC systems.
 
Wow... if that's true then that's really good news!

However, I can't help but think that benchmarking is even less useful than usual this time around, given the very skittish benchmarking results on that page (and also with some of the other reviews I've seen). I think it's apparent that there's a bit of unevenness between each test in a suite (e.g. OpenGL vs. Quartz), and between PPC/Intel. Or maybe OS X on Intel just needs some polishing still. It's the real-world stuff that counts and from what posts and reviews I've read, the feedback is pretty good, especially in the ever-important Snappy UI™ department.

Thanks for pointing out that macintouch page... now all I need is a full review from BareFeats to feel confident in getting an iMac (well that... and selling my PowerBook😉)
 
Given this guy actually used a benchmarking tool specifically designed for Java, I would actually tend to believe it - although I don't necessarily think that Applet performance is reflective of the entire JVM but you can't argue with the result... (http://www.netlib.org/benchmark/linpackjava/)

Lee Givens (Macintouch link above) said:
Hello All, During my post-Macworld travels, I've been anxiously awaiting my new MacBook Pro and searching Apple Stores far and wide for iMac Core Duos to run tests on (ones I forgot while at Macworld). The latest test done at the Los Angeles South Coast Plaza Apple Store on 1/18/06:

222 Mflop/s; Other Mac OS iMac Core Duo/2x2Ghz/10.4.4/1GB RAM/Safari/J2SE1.5; 1/18/06 Lee Givens

You can view other timings and you'll see that the iMac Core Duo is equal to a Dual 1.8/2.0 G5 I tested a while back. (Linpack timings: http://www.netlib.org/benchmark/linpackjava/timings_list.html)

I'm writing this on my PowerBook G4 (linpack ~54 Mflops/s). If you're wondering about performance and confused by XBench, feel comfortable that the MacBook Pro should hit well over 200 (I averaged 202 on 4 runs on the Macworld booth MacBook pro). I only feel it will get better as Apple optimizes for Intel.

If you run Java apps, you'll see a 3.5x performance boost on the new Intel Core Duo machines. Finally, real Java performance.
Click to expand...
 
I agree... those results look pretty impressive. Here's hoping the speed increase is realized in jEdit... code completion, at lightning speed! 😎
 
dolphin842 said:
I agree... those results look pretty impressive. Here's hoping the speed increase is realized in jEdit... code completion, at lightning speed! 😎
Click to expand...

For comparison, I just ran on my Windows machine (P4 HT - 1GB RAM) in FF1.5 and averaged roughly 140 Mflops/s +- 2. So if the intel macs are getting 200+, thats AWESOME! I really need to test one out.
 
I was curious and checked out the test as well. Given that I score around a 50 on my PowerBook, the iMac would be quite an improvement!
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back