PDA

View Full Version : Cinebench R10 beta !!!


jerry0520a
04-18-2007, 01:50 PM
http://www.dailytech.com/Intel+Unveils+Penryn+Performance/article6958.htm

e[dub]
04-18-2007, 05:23 PM
interesting that the quad core penryn scores 1.7 times higher than dual core, using CB R9.5, but 1.85 faster using CB R10b...

I'm no hardware wiz, but I assume that means that CB R10 is optimized for the Intel chips? Is that reflected in the current build of Cinema?

Shademaster
04-19-2007, 10:48 AM
']interesting that the quad core penryn scores 1.7 times higher than dual core, using CB R9.5, but 1.85 faster using CB R10b...

I'm no hardware wiz, but I assume that means that CB R10 is optimized for the Intel chips? Is that reflected in the current build of Cinema?

I believe updated to run more efficient on MORE cores instead of just Intel :).

Simon Wicker
04-19-2007, 01:16 PM
you can't compare different versions of cinebench, only different machines running the SAME version (so all 9.5 v.s. 10.1 cinebench comparisons are not valid).

cheers, simon w.

kossoolli
04-20-2007, 02:11 PM
Heres a screen of the beta:
http://www.gamestar.de/_misc/img/detail.cfm?pk=63558&fk=1469580&id=EL_1176941819849629060976

e[dub]
04-20-2007, 03:49 PM
looks NICE!

ooo
04-20-2007, 04:07 PM
I bet Srek is riding on one of those :cool:

odo

cmyk
04-21-2007, 07:21 PM
Anyone run the beta on an 8-core Mac Pro yet? I just ordered mine today and am dying to know how it compares...

-kev

e[dub]
04-21-2007, 09:44 PM
where would one get the beta? is it available for download?

Malameel
04-21-2007, 11:19 PM
I am interested in the new Cinebench. Not that we make movies of numbers scores but it is fun to run everytime you get a faster computer.
M

Srek
04-22-2007, 06:38 AM
The new MAXON CINEBENCH R10 will not become publicaly available as a beta. You will have to wait for the release of the final version. Shouldn't take to long though.
Cheers
Björn

rob rhodes
04-23-2007, 12:35 PM
Hang on so if cinebench 10 is more optimised then does that mean C4D rendering will become more optimised? All seems a bit crazy then this benchmark thing, if CPU efficiency is based on your render engine efficiency. So has cinebench been lying (too strong but...) to us if we test our PC with 9.5 it is less efficient than with 10. So now the Mac Pros will look more efficient and the intel chips scaling better? I just thought cinebench was a measure of how fast/efficient your machine was regardless of software used. Sorry for the ramble I might have got something wrong somewhere, please feel free to clear this up.

LucentDreams
04-23-2007, 01:18 PM
as computers evolve so must benchmark tools. Benchmarks use a certain level as their base and compare all results according to that base as new systems and technologies come out and speets hit new levels then yes benchmarks start to become less accurate, this is why there is a new version of sysmark every year.

With the current cinebench the harshest fact is that some systems render so fast that the rendering benchmark doesn't even have enough time to measure a suitable result. Optimzed doesn't always mean its faster, optimal means better, most favorable or desirable. By saying its optimized it simply means its improved to provide the most ideal results possible for the latest in system changes.

rob rhodes
04-23-2007, 02:10 PM
OK cheers Kai, so the 4 / 8 core machines are actually more effiecient than we have previously given them credit for? It has been the benchmarking software that has been holding them back, rather than their poor scaling performance? Something like that?

Srek
04-23-2007, 02:14 PM
I just thought cinebench was a measure of how fast/efficient your machine was regardless of software used.
Nope, that's simply not possible, except if you count synthetic tests and those have the problem that the results are usualy not transferable to real world applications and therefore only helpfull for marketing or scientific uses.
CINEBENCH is using CINEMA 4D functionality for benchmarking so it primarily tests how your system will perform with CINEMA 4D. However, due to how CINEMA 4D is implemented the results become usefull even if you don't intend to use CINEMA 4D but another application with similar functionality. The less similar your intended application is to CINEMA 4D the less reliable the results become (i wouldn't use CINEBENCH to benchmark a machine for use as a web or database server).
Short example: If you use CINEBENCH to test if your new system will be faster for videoediting the results will very likely not be helpfull. Videoediting relies much more on memory and disk IO speed than rendering does.
CINEBENCH gives usefull results if you intend to test for rendering and modelling speed.

Cheers
Björn

Srek
04-23-2007, 02:18 PM
OK cheers Kai, so the 4 / 8 core machines are actually more effiecient than we have previously given them credit for? It has been the benchmarking software that has been holding them back, rather than their poor scaling performance? Something like that?
It's not that simple. The over all performance (especialy for multithreaded applications) depends on many factors, the most important are the code optimization of the application, the development tools, the hardware and the OS. Any change to one of these factors and over all efficency changes. To make sure new development in regard to operating systems, hardware and development tools is taken into account not only CINEMA 4D is adpated to them but at the same time CINEBENCH.
Cheers
Björn

CGTalk Moderation
04-23-2007, 02:18 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.