L2 cache, 512KB vs 1MB,E1200 vs E2140

*snip*
posts gone
*snip*
i should have started this when i first did the comparsion... the basis for my reponse in the thread a = loss. 4mb vs 2mb/2mb vs 1mb/1mb vs 512kb, doing the number crunching. the losses are around the same % with a slight gain each time. If you lose the same % every time from a 50% reduction in L2. will you really notice that 10% or even cumiltive 12-14% final loss, when comparing the 2 cpus?

This is for others to join in, hopfully.. this is not to be a flamebait at all... just tring to get a discussion on it...
-----------------------------------------------------------------------------
this is where i leave off ..........

Kuroimaho said:

Did you bother reading the links I posted ?
I copy them once more just in case.

Celeron [email protected] VS [email protected]

Click to expand...

i did read those and i didnt see the [email protected] last night.. call it me being tired or what ever, i really dont care. what im comparing like in my OP about this was/is high res gaming,50% L2 loss. which leads me to ask you then, do you game at 1024x768 on your 2900XT/30in display?

Kuroimaho said:

Okay, so if the bench I posted isn't good enough for you again, care to show us where did you get your information proving that cache does not matter in games ? That must be more detailed for sure.

I guess you won't like this either because they came to the conclusion cache matters. Does cache size really improve performance ?

Click to expand...

i could ask if you read what i posted or even what you posted/linked to?

Evilsizer said:

I said L2 had little if any impact on gaming comparing a cpu with double the L2. *added* meaning 1mb down to 512kb or using the simular like i did before 2mb down to 1mb. since it is still the same % lost.

Click to expand...

as well even i noticed this
http://xbitlabs.com/articles/cpu/display/celeron-e1200_5.html#sect0 <-- testbed using a NV 8800GTX
http://www.xbitlabs.com/articles/cpu/display/pentium-e2160_7.html#sect0 <-testbed using a ATI x1900XTX
when reading the whole review not just the benchmarks, from your links.

you expect me to use 2 testbeds with different video cards. to compare L2 cache sizes in gaming for this debate?

as well that last THG link you provide does the following for their conclusion. that link is comparing a 1mb to 4mb, they are basing there conclusion on that. they even said it too.

THG said:

If you compare the benchmark results of the 3D shooters Prey and Quake 4 with typical gaming benchmark results in our CPU Charts, the performance difference of 1 MB vs. 4 MB L2 cache roughly equals one clock speed increment. The same applies to the video transcoding benchmarks for the DivX 6.6 and XviD 1.1.2 codecs and file compression using WinRAR 3.7. The CPU-intensive benchmarks of 3DStudio Max 8, the Lame MP3 Encoder or the H.264 Encoder V2 by MainConcept don't benefit much from increasing cache sizes, though.

Click to expand...

im have been basing my conclusion on the fact at with a 50% loss in L2 gaming in not that effected. im not basing my conclusion like those reviewers are looking at 512kb/1mb to 4mb difference. i cant say that enough.... even if we just look at the 512kb to 1mb results. you assume they are drawing there conclusion on this, when they are not.

even if you stick to the 1024x768 results alone. do you really notice a 20 or even 40fps loss in fps. when your well over 100? to futher assume from the links since no mention is said. lets just say the fps numbers i put up are MIN FPS,not max or avg.

I have to agree that reviews are not 100% accurate but in order for us to really know if the difference is huge, someone needs to have both these chips and test it out.

o trust you and me thats coming!! that was one of my big points from the other thread.. i will do a L2 comparsion each cpu at the same speed(s). I have 3 cpus to work with just need 2 more. one being the E1200 and the other being a E6420. as all of the cpus i have are 8x multi's. might as well keep it the same thru out, not to metion used E6420 are cheaper then E6600's.

Another thing i have been keeping in my mind. is if someone is on a set $$ for a rig, why not a E1200 vs E2140(both max out at 3.2-3.4ghz)? what if the $16(according to prices on newegg) difference, is the difference between a good gpu and a bad one. what if its the difference between getting a WD raptor and not getting a raptor? where for a gaming rig, both HD/GPU are much bigger factors then L2 size.

from your other comment though AO ,
http://xbitlabs.com/articles/cpu/display/celeron-e1200_8.html
http://www.xbitlabs.com/articles/cpu/display/pentium-e2160_8.html#sect1
Knowing that in 3DMARK 06, the video card plays no role in CPU score.
3DMARK06 E1200 CPU SCORE = 1365
3DMARK06 E2140 CPU SCORE = 1372
E1200 is only .6% difference per 50% less L2 then the E2140 in that benchmark. thats not that big of a difference... you cant use the other score cause well look at it... the E1200 is showing 6528 vs 5500 for the E2140, thats just showing the difference in gpu's. which herin goes to the other factors im keeping in mind. GPU is more a factor then L2 for the same cpu speed.

*this took a while to type up... gonna have to start doing this, to get my WHOLE thought out at once. not in pieces like it has been coming out..