It's a shame that wide dissemination of these tests is going to completely destroy PM sales (and hence Apple's cashflow/profitability) until those 970 boxes are out of the door.
More so than they are now? PM sales have already tanked. People only want to upgrade to the same hardware a limited number times, especially when the same hardware continues to fall behind the competition.
When I say "completely destroy", I mean utter flatline: I have seen a few folks purchasing PMs recently, but this could sound the death-knell even for that trickle.
It's too bad the article makes no mention of the graphics cards used, or anything other than it's a beta Panther, and those #'s. I'd like to know exactly what we are dealing with here.
I posted a new thread since I didn't see this news mentioned in the two or three other active 970 threads...at least when I looked at the posts dated May 5th earlier today.
People. I just mailed Lionel of Macbidoullie about the Bryce benchmarks. Here's the answer I got:
Quote:
It seems there was a problem with this bench. However I didn't want to modify my source's information to stick to reality. The results are the ones sent to me.
I don't care the Bryce benchmark, but I DO care about Cinema
The Cinema benches don't look right either. The dual 1.8 970 should have a render time of 10-13 seconds, not 18 (based on the single 1.4 970 score of 29). Something's wrong here.
The Cinema benches don't look right either. The dual 1.8 970 should have a render time of 10-13 seconds, not 18 (based on the single 1.4 970 score of 29). Something's wrong here.
Some scene files also affect rendering speed.....
As far as I can tell. Rendering in Cinema is divided into two parts. Preparation and Rendering. Each frame has to be prepared before rendering. The preparation part is NOT MP aware. If there are lots of texture maps and shadows and lights......there will be a huuuuuuuuge slow down in the preparation before it actually starts the rendering
The Cinema benches don't look right either. The dual 1.8 970 should have a render time of 10-13 seconds, not 18 (based on the single 1.4 970 score of 29). Something's wrong here.
based on what logic?
remember there are diminishing returns to multiple processors, especially with apps which are not explicitly optimized for them...
Cinema gets ~1.85x improvement with dual processors (according to CineBench). So 29 sec. for single 1.4 970 would translate to ~15.6 seconds for a dual 1.4. A dual 1.8 970 should score better than 15.6 sec. (I did a rough horseback guesstimate to get 10-13 sec.)
edit: Cinema is most definitely optimized for dual. It's even optimized for hyper threading where supported (some of the Pentiums?)
just one other comment: shouldn't a single 1.8 970 be ~25% (or a bit more) faster than a single 1.4 970? I'm not a programmer or a math whiz by any means, but this seems right. So a single 1.8 ought to bench at 21 sec or so, so a dual 1.8 ought to bench at 11 - 13. I dunno, I just make pictures
Cinema gets ~1.85x improvement with dual processors (according to CineBench). So 29 sec. for single 1.4 970 would translate to ~15.6 seconds for a dual 1.4. A dual 1.8 970 should score better than 15.6 sec. (I did a rough horseback guesstimate to get 10-13 sec.)
edit: Cinema is most definitely optimized for dual. It's even optimized for hyper threading where supported (some of the Pentiums?)
Please read my response above regarding preparation and rendering. Dual CPU system on short render isn't very efficient compare to huge rendering that takes at least 2-3 minutes.....
Or maybe those people accidentally swapped the Cinema and Bryce benchmarks
Comments
I can't stand what you put me through.
Your life's a lie, that you hide.
Is it that terrible being you inside?
I can't stand, oh, the thought of you.
I can't stand all the things you do.
What do you try to justify?
You were just too scared to be you inside.
Let it all out...
Let it all go...
I look at you, all I see, is a man too afraid to really be...
I can't stand what you put me through.
I can't stand even the thought of you.
Your secret lies that you hide.
Is it that terrible being you inside?
You try so hard to be wanted.
False emotions tells you fronted.
I think being a person relies on one thing:
Be yourself, let you come through.
You're too afraid to really be,
Someone who isn't false, who doesn't care to be.
Be yourself, let you come through!
Fake!
You'll regret it, you'll regret it...
Sh!+ meet pants!
PLEASE LET THIS BE TRUE
Originally posted by Overhope
It's a shame that wide dissemination of these tests is going to completely destroy PM sales (and hence Apple's cashflow/profitability) until those 970 boxes are out of the door.
More so than they are now? PM sales have already tanked. People only want to upgrade to the same hardware a limited number times, especially when the same hardware continues to fall behind the competition.
Originally posted by onlooker
It's too bad the article makes no mention of the graphics cards used.....
Rage 128 with 8MB VRAM
Originally posted by LudwigVan
Apologies for the disorientation.
I posted a new thread since I didn't see this news mentioned in the two or three other active 970 threads...at least when I looked at the posts dated May 5th earlier today.
No hard feelings
Mac fans, our wait will be rewarded. The fight is over and Apple will soon rule the world !
Oh, and...
"There are no American infidels in Baghdad. Never!"
Originally posted by NMR Guy
Oh, and...
"There are no American infidels in Baghdad. Never!"
It seems there was a problem with this bench. However I didn't want to modify my source's information to stick to reality. The results are the ones sent to me.
Perhaps Dual 970 Works like a dual core CPU.
If it was a fake, i didn?t make a so big mistake
Best regards
Lionel
I don't care the Bryce benchmark, but I DO care about Cinema
The Cinema benches don't look right either. The dual 1.8 970 should have a render time of 10-13 seconds, not 18 (based on the single 1.4 970 score of 29). Something's wrong here.
Originally posted by artcat
The Cinema benches don't look right either. The dual 1.8 970 should have a render time of 10-13 seconds, not 18 (based on the single 1.4 970 score of 29). Something's wrong here.
Some scene files also affect rendering speed.....
As far as I can tell. Rendering in Cinema is divided into two parts. Preparation and Rendering. Each frame has to be prepared before rendering. The preparation part is NOT MP aware. If there are lots of texture maps and shadows and lights......there will be a huuuuuuuuge slow down in the preparation before it actually starts the rendering
Originally posted by artcat
The Cinema benches don't look right either. The dual 1.8 970 should have a render time of 10-13 seconds, not 18 (based on the single 1.4 970 score of 29). Something's wrong here.
based on what logic?
remember there are diminishing returns to multiple processors, especially with apps which are not explicitly optimized for them...
based on what logic?
Something like this:
Cinema gets ~1.85x improvement with dual processors (according to CineBench). So 29 sec. for single 1.4 970 would translate to ~15.6 seconds for a dual 1.4. A dual 1.8 970 should score better than 15.6 sec. (I did a rough horseback guesstimate to get 10-13 sec.)
edit: Cinema is most definitely optimized for dual. It's even optimized for hyper threading where supported (some of the Pentiums?)
Originally posted by artcat
Something like this:
Cinema gets ~1.85x improvement with dual processors (according to CineBench). So 29 sec. for single 1.4 970 would translate to ~15.6 seconds for a dual 1.4. A dual 1.8 970 should score better than 15.6 sec. (I did a rough horseback guesstimate to get 10-13 sec.)
edit: Cinema is most definitely optimized for dual. It's even optimized for hyper threading where supported (some of the Pentiums?)
Please read my response above regarding preparation and rendering. Dual CPU system on short render isn't very efficient compare to huge rendering that takes at least 2-3 minutes.....
Or maybe those people accidentally swapped the Cinema and Bryce benchmarks
Originally posted by Marcus
hehe this has just made the front page of /.
And, between this and the Music Store numbers today, Apple's stock was up over $2 at some points today.