PDA

View Full Version : amd in relevance to intel. a little help please.



noktekniq
06-22-2006, 09:48 AM
so i have been an AMD guy ever since. but then again when conroe comes out i'm heading straight to intel if the price is promising and the oc guarentee is promising unlike amd where you have to buy good batches for high clocks.

so here is the question. how does it work?
AMD opteron 170 at 2.75ghz. how fast is that in relevance to a intel speed chip? speed wise that is.
so like AMD 4000+ is that like a intel at 4.0ghz? or not? i dont' get it?

cupholder2.0
06-22-2006, 10:15 AM
so i have been an AMD guy ever since. but then again when conroe comes out i'm heading straight to intel if the price is promising and the oc guarentee is promising unlike amd where you have to buy good batches for high clocks.

so here is the question. how does it work?
AMD opteron 170 at 2.75ghz. how fast is that in relevance to a intel speed chip? speed wise that is.
so like AMD 4000+ is that like a intel at 4.0ghz? or not? i dont' get it?

Which Intel chip? If its a conroe, a 1.63 Ghz Conroe would probably own the AMD 4000+ easy. If you were comparing to a P4, it would probably be 3.4-3.6 Ghz.

Mikesnav
06-22-2006, 10:48 AM
2.8GHz Opteron... I'd say a 4GHz+ Pentium.

Conroe... ~1.8GHz

And yes, AMD's single core PR is a somewhat decent way to compare performance in regards to the single-core Pentium's frequency.

viccyran
06-22-2006, 11:53 AM
Their PR's are relative to their own speeds from what I heard.

As in a 4000+ is equivalent to a barton running at 4ghz? Something of the sort.

MRBIGSHOT
06-22-2006, 12:58 PM
amds naming sceme has always based apon comparitive speed versus a Thunderbird core athlon cpu. they started this when they first started using the athlon XP name, that coensided with the launch of the Thourobred A cpus. against popular belief it was not comparing to intels pentium 4 offerings, although at the time of the intel northwood A cores both processors performed nearly identical. the naming sceme became a muddy mess with the launch of the 754 athlon64's as all kinds of numbers were being thrown around. pr changes due to on die cache and clockspeed's confused everyone as in most applications the added cache did not improve performance and certinly not to the tune of the 200mhz clock speed gap.

now the new naming sceme
if the processor has an E prefix you will know that it uses more that 50 watts
if the processor has a T prefix it will use less than 24-49 watts (mobiles)
if the processor has a X prefix it will be the xtreme edition cpu

the 6xxx denotes a core duo conroe cpu
the 4xxx denotes a core single conroe cpu
the 2xxx denotes a merom duel core cpu
the 1xxx denotes a merom single core cpu

intels old naming sceme is not to diffacult to understand once you figure out what each number means, ie the first number is the processor type.
3xx was the celeron line
5xx(j) were the single core prescott 1 meg cache - (j) dented the 64 ability
6xx was the single core 2 meg cache prescott cores
7xx was the dothan core moble cpu's
8xx was the smithfield
9xx was the pressler cores

andyisc00l
06-22-2006, 07:12 PM
Sometimes I think people here overestimate conroe and are confused by the results they see that are posted here...conroe @ 1.8ghz will probably not beat a amd @ 2.7ghz, or at least thats the looks of it..personally I don't think we have a single web page with a credible conroe/amd comparison..not going by super pi times or 3dmark there aren't too many credible reviews..it's definatly more powerfull clock per clock but 1.8ghz to 2.7ghz seems pretty steep.

From that logic will we say a conroe e6600 overclocked to 3.6ghz will beat an amd dual core @ 5.4ghz..and a 5.5ghz conroe beating an 8.3ghz amd?? Not to pick on you many people have said similar numbers and I'm yet to be convinced..:-/

[XC] DragonOrta
06-22-2006, 09:16 PM
I'm not so sure of the 2.8ghz Opteron being the same speed as a 1.8ghz Conroe, that seems a little off, but Conroe improves in every way against just about any single CPU.

grimREEFER
06-22-2006, 09:48 PM
wait, a 1.8ghz 2mb l2 cache holds its own against an fx-62 , so, yea, it seems ok