Dual Core Vs. Single Core

Become a member of the CGSociety

Connect, Share, and Learn with our Large Growing CG Art Community. It's Free!

Thread Tools Search this Thread Display Modes
Old 08 August 2007   #1
Dual Core Vs. Single Core

Hey all. I'm currently doing some googlin' to find an answer but will post here as well in the event of getting a response quicker.

I'm looking for a comparison or equivalency chart of dual core versus single core processors. For example, what would a single core 3.5 GHz equate to in a dual core set up? Is there a general rule of thumb for the performance factor? I came across something that mentions a dual 2.8 will be as a fast as a single 3.8.

I'm new to this whole dual/quad core scene and want to have some sort of baseline when building a new system based on particular software hardware requirements.

Hope this makes sense and thanks in advance for reading and possibly responding.

Listen: Nick Cave & The Bad Seeds
Old 08 August 2007   #2
1.8, give or take, is a decent way to measure it. so a dual-core processor will render a scene 1.8 times faster - roughly spoeaking anwyays, it's depends on the program. Sunsequently, a quad core is roughly 1.8 times faster than a dual-core.

However, this speed difference doesn't apply across the board. 1.8 is a good number for rendering, which is very optimised for multiple threads.

Oblivion (a computer game) see's about a 20% improvement with the addition of a second core, I havn't seen anything about quad, but guessing it's 0. That game would see greater benefit for a faster single core, rather then two slower dual-cores.

EDIT: That 1.8 rule levels out at about 4, most especially for Intel cpu's, where after 4, the FSB starts to beocme a real bottleneck - AMD's scale a lot better, but still, as more core's are added, the overall benefit starts to decrease.
Old 08 August 2007   #3
I've just been playing with an 8 core machine, and running multi-process tasks (such as bucket renders), it was a bit less than 6 times as fast on multi-core as it was using a single core.

The biggest thing I've found is that software generally doesn't support your extra cores - so you only end up with one chip doing all the work. However, if your OS deals with it sensibly it will throw all the other tasks to the other chip so your computer won't grind to a halt whilst once of the chips is being hammered.

It's not just limitations of how the chips work, but how the software works, some stuff, just doesn't work well in a multi-thread environment, and you're just better off with a bigger faster CPU.
Get off my lawn.

Twitter: salmonmoose
Old 08 August 2007   #4
1.8 is only correct if the chips share the same basic technology and clockrate.
Every comparison you make has to be based on a common factor. Clockrate, number of cores, size of cache etc. do not qualify for this, since they are vastly different between chips.
On your original posting, you want to compare a single core 3.5 GHz processor with a dual core 2.8. Well, the comparison can only be made by using a benchmark with the exact CPUs you are looking at. 3.5 GHz are not available for any CPU with the same basic technology as the 2.8 GHz Dual Core CPUs use. If you only go by GHz one would expect the 3.5 GHz CPU to have a speedadvantage of 1.25 on single threaded tasks. In reality the 2.8 GHz will more likel by 1.5 times faster or even 2.7 times faster when it comes to multithreaded stuff. This is based on my experience with those CPUs, not measurements.
Maybe take a look at CINEBENCH (or any other 3D rendering benchmark) for results. The table of CINEBENCH results from Mash should give you an idea on what to expect.
- www.bonkers.de -
The views expressed on this post are my personal opinions and do not represent the views of my employer.
Old 08 August 2007   #5
Srek is right. This 1.8 factor really only applies to applications that can utilize multiple cores/CPUs AND between CPUs that are based on the same design.

For example, a Pentium D vs a Pentium 4 would result in a 1.8x speed up ONLY in multi threaded apps. And only with all other factors identical (Clock speed, cache size, etc)

If you are trying to compare the Pentium 4 to a Core 2 Duo, the underlying technology is vastly different. Even a single core of the Core 2 Duo will most likely out perform a Pentium 4 of more than 1GHz faster. The design of the individual cores is much faster.

Basically, you can only compare between CPUs like this if all other factors are the same. If they're not, then your comparison is not valid as other factors can contribute to the speed difference.

In either case, if you are considering a single core Pentium 4 for anything, forget it and get a Core 2 Quad :P
Dual Opteron 246 | Tyan Thunder K8WE | 4GB RAM | Geforce 7800GT
Checkout my contribution to the LW 9.2 Demo Reel:
Old 08 August 2007   #6
Thanks for all the great feedback.

Here's another example of what I'm looking for. Let's say a game lists the recommended cpu requirement as a P4 3.5 GHz or equivalent. What would be the minimum dual-core cpu speed be to sufficiently handle that? Would I look at getting a dual-core 1.6, 2.0, 2.4, or....? Could I say a 2.0 (x 1.8 factor = 3.6) dual-core would be roughly comparible?

I'm looking at buying either a quad-core desktop or duo-core laptop and want to be sure it will run the 'latest & greatest' apps based on the cpu I choose.

Listen: Nick Cave & The Bad Seeds
Old 08 August 2007   #7
You may find this article useful: http://www.tomshardware.com/2007/07...2007/index.html
Old 08 August 2007   #8
A 2.4 GHz Core 2 Duo or Core 2 Quad will be about as fast as a 3.5 GHz P4 for single task stuff. Sadly games are only starting to make use of multiple cores so any core more than one is of very limited use most of the time. Regardless of how bad the game supports multithreading you still get an advantage though. All OS task as well as Antivirus and whatever will not slow the game done as much as it would on a single core system.
- www.bonkers.de -
The views expressed on this post are my personal opinions and do not represent the views of my employer.
Old 08 August 2007   #9
Thanks for the link Adam.

And thanks Srek for the clarification. That helps.


Listen: Nick Cave & The Bad Seeds
Old 08 August 2007   #10
Thread automatically closed

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.
CGTalk Policy/Legalities
Note that as CGTalk Members, you agree to the terms and conditions of using this website.
Thread Closed share thread

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Society of Digital Artists

Powered by vBulletin
Copyright ©2000 - 2006,
Jelsoft Enterprises Ltd.
Minimize Ads
Forum Jump

All times are GMT. The time now is 01:57 AM.

Powered by vBulletin
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.