Welcome Guest ( Log In | Register )

Bump Topic Topic Closed RSS Feed

Outline · [ Standard ] · Linear+

 Lenovo Y580 on hand, 9 Aug.295.93 so far the best driver Y580

views
     
skeith
post Jun 14 2012, 09:24 AM

Casual
***
Junior Member
342 posts

Joined: Jul 2007
sweet machine bebe!!!

It came out too late, so i get my clevo 1st!

orelse, tis model will definitely get into my consideration
skeith
post Jun 14 2012, 09:38 AM

Casual
***
Junior Member
342 posts

Joined: Jul 2007
QUOTE(PF T.J. @ Jun 14 2012, 09:34 AM)
Your Sager is way better bro  thumbup.gif
*
I pay the price man rclxub.gif

and GTX660 with no more then 3400 is tempting

This post has been edited by skeith: Jun 14 2012, 09:39 AM
skeith
post Jun 28 2012, 10:26 AM

Casual
***
Junior Member
342 posts

Joined: Jul 2007
just realized ur GTX660M clock speed is so high! O.O

my 670M just 598MHz (core clock) and 750MHz (memory clock)

why?
skeith
post Jun 28 2012, 01:01 PM

Casual
***
Junior Member
342 posts

Joined: Jul 2007
QUOTE(Zepx @ Jun 28 2012, 12:47 PM)
Because 660M is Kepler and 670M is Fermi....
*
so what u trying to say? more CUDA in GTX660M, more clock speeds in GTX660M but less power then GTX 670M.

are u trying to say kepler sux?
skeith
post Jun 28 2012, 07:52 PM

Casual
***
Junior Member
342 posts

Joined: Jul 2007
QUOTE(Acid_RuleZz @ Jun 28 2012, 01:49 PM)
Totally different architecture, you can't compare them core to core imo. hmm.gif
*
i know its different architecture. but from the basic spec, it is somehow shows that kepler or GTX660M itself use more resource, but giving less performance then fermi GTX670M. hmm.gif
skeith
post Jun 28 2012, 09:13 PM

Casual
***
Junior Member
342 posts

Joined: Jul 2007
QUOTE(Acid_RuleZz @ Jun 28 2012, 09:03 PM)
Resource? How about, Fermi use alot more power than Kepler but its not that much faster.

and again.
*
true oso lar... but just feels something not right.

have to do more research on kepler architecture when i'm free.
skeith
post Jun 28 2012, 10:11 PM

Casual
***
Junior Member
342 posts

Joined: Jul 2007
QUOTE(Acid_RuleZz @ Jun 28 2012, 09:24 PM)
Maybe this can shed some light. tongue.gif
*
good! light and simple comparison. less power to move more cores, much lesser.


Added on June 28, 2012, 10:12 pm
QUOTE(Zepx @ Jun 28 2012, 09:47 PM)
The key is that you can't really compare two different architectures just on core clock speed... in that case we could easily compare AMD and Nvidia just by the core speed, which would generally be different because of how scarce components can be placed in the die. What Nvidia has removed from the Fermi has to be compensated by something in the Kepler architecture, so it's no big deal if the clock speed increases, but the result of the entire die does not beat the 670M. Again Kepler is a focus on efficiency and a slight improvement on performance, in short a give and take. Having best of both world is what we called the ideal GPU.

If you want to compare clock speed, you need to make sure other things like architecture are constants in the discussion subject.
*
true. wat u said was right. its hard to compare two performance with different architecture. everything had their pros n cons, and different focus point.

This post has been edited by skeith: Jun 28 2012, 10:12 PM

Topic ClosedOptions
 

Change to:
| Lo-Fi Version
0.0287sec    0.75    7 queries    GZIP Disabled
Time is now: 1st December 2025 - 06:26 AM