Looks like Conroe ES are spreading. Victor Wang our own XS XIP introduced first initial results with CoreDuo E6600. Using upcoming mainstream i965 based mainboard and without any overclocking Super PI 1M test passed in just 21 seconds.

www.xtremesystems.org/fo...read.php?t=95021

whats a XS XIP?

xtreme systems xtremly important person? lol

what would a fx-60 get on this test? Super PI 1M?

dam its fast kicks the shit out of a FX55. but its intel.

the thing i like is that the benchies coming out are based off the basic ones... i'm still saving for the 3.33 XE 1333fsb =)

Intel’s Conroe Acquired via Ebay Auction

www.xbitlabs.com/news/cp...60402235626.html

LOL..... lucky guy =)

If anybody missed it, there is an interesting technical article here arstechnica.com/articles...dia/cpu/core.ars about the internals of the new Core generation CPUs.

I was putting off upgrading my home system until AM2 was available, but these Core cpus are starting to look very good.

Could anyone please link me to a map or descriptions of the conroe's?

Also, how do they compare to the current AMD range, for price / performance, are they dual core?

don't go there =P

go here instead LOL

forum-replies.cfm?t=452582

there's alot of link in there that will answer all your question... benchies/specs/so on and/soforth =P

Viper6 writes...

without any overclocking Super PI 1M test passed in just 21 seconds.

omg...

My P4 @ 4Ghz can only get ~32sec..

Just shows how much better these Conroe's are.

i want conroe! >___<

me 2 =D
4mb of cache , 1066fsb

i wished they come with stock water cooling =P (ocing .....)

((michael)) writes...

but its intel.

and 2 years down the track AMD will come out with something slightly faster and you'll say "but its AMD."

Davodude writes...

4mb of cache , 1066fsb

remember the XE model has 1333fsb and 4mb cache =) @ 3.33ghz.... oooooh mama

Benly writes...

omg...

My P4 @ 4Ghz can only get ~32sec..

Time for an upgrade soon I think, it won't be intel though. :P

My old 2100+ XP B core @ 2278mhz (217x10.5) takes exactly 44.672 seconds.
@ 2300mhz (200 x 11.5) takes exactly 50seconds.

Still not bad for an old clunker. :P

About time, so glad I didn't venture over to the dark side.

Wow, that is a seriously quick 1m super pi time.....

It will be interesting to see AMD's responce to this......I am sure they will produce an equally amazing product, just that they have not released test versions of it, for publicity....

does anybody know how much these chips will be when they come out, as in the midrange chips. WIll they be comparable to the chips already out? As in around $400 for a decent midrange, or will they be expensive and later come down?

Simon L writes...

It will be interesting to see AMD's responce to this......I am sure they will produce an equally amazing product,

yeah not till mid next year though they reckon... so you amd fanois have got some time under your belt =)

ministryofsound writes...

so glad I didn't venture over to the dark side.

meaning amd or intel?

However, once you go over the 4MB limit, Conroe is slower than Athlon 64 at the same clock. Both the Cryptography and STREM tests use a lot more than 4MB, larger than Conroe's 4MB cache, and Conroe immediately falls below Athlon 64 on the performance curve

sharikou.blogspot.com/20...eing-busted.html

Interesting times ahead :P

I dunno, i mean a Yonah can keep up with the best A64 has to offer, i dont see why the conroe wont do better. We will just have to wait and see, maybe he didnt have correct BIOS, or thats an engineering sample (im not aware if it was mentioned in the article)

Viper6 writes...

Interesting times ahead :P

The following results were obtained by running 32 bit ScienceMark binaries optimized for Intel Pentium:

It was optimised for a 6 generation old platform? (P-M, P4,P3,P-Pro,P2,P1)

is ScienceMark optimised for dual core? Otherwise write of the whole test.

Basically they go onto say that the 4MB Cache gives the Conroe a performance boost, logic tells you this, but after that teh A64 is acautally faster clock to clock (since when has intel been on par clock to clock?).

Most modern applications need a lot more then 4MB. IE needs at least 50MB when viewing a normal web page(with Flash, JS, DHTML, AJAX..); Photo Editing apps need around 40MB; FireFox takes 23MB when I use it to view yahoo.com; DivX grabs 23MB even before I open a video...

But every L2 cache hit is much faster then a RAM hit... With the Conroe having more L2 cache then it has more cache hits and would therfore get a speed increase.

The tests werent even done by the guy posting them... how sad.

Ive also said it before: dont believe any tests that a company does aganist it's competitor.

EDIT: RAM in the L2 cache bit was L3 before, but L3 was the incorrect term.

Hellman109 writes...

But every L2 cache hit is much faster then a L3 cache hit... With the Conroe having more L2 cache then it has more cache hits and would therfore get a speed increase.


are u sure about this? i thought lower level cache tend to have higher associativity and thus higher hit times. the higher level cache is purely there to minisize miss rate.

so conroe has level 3 cache? edit: <<- nrv. just realised it only have up to L2

Viper6 writes...

Interesting times ahead :P

You trust information on Blogspot? Sheesh!

Even so, how many standard apps that Whirlpoolians would use (games, video encoding etc etc etc) would have data sets anywhere near 4Mb? I'd wager none.

(Remember that total memory usage of software is not a useful comparison for what would normally need to reside in cache for decent performance)

Squallff8aus writes...

so conroe has level 3 cache?

Oops... my mistake, I mean RAM, Ill edit in a second.

the higher level cache is purely there to minisize miss rate.

That is the point of higher cache rates.

are u sure about this? i thought lower level cache tend to have higher associativity and thus higher hit times.

The lower the cache the more it has to run through a replacement policy and such, and the higher rate you have of data that was cached, but was replaced for what it is working on at the time.

Ofcourse the first reference you have a compulsory miss, but after that the more hits you have, the less replacements, so the more efficient the system runs.

Beefy writes...

Even so, how many standard apps that Whirlpoolians would use (games, video encoding etc etc etc) would have data sets anywhere near 4Mb?

this isn't ram though. cache structure matters to the particular architecture not to the specific s/w. thats the idea anyway. Its suppose to be as s/w independent as possible.

Hellman109 writes...

The lower the cache the more it has to run through a replacement policy and such

thats a nice way of putting it. i haven't thought of it that way before :)

Squallff8aus writes...

cache structure matters to the particular architecture not to the specific s/w. thats the idea anyway. Its suppose to be as s/w independent as possible.

That's not really what I'm considering.

The author of that blog is trying to suggest that any software with a RAM usage of more then 4Mb is going to be slow on Conroe. I'm saying that this is a false assumption, because RAM usage is not related to what needs to be in the L2 cache.

For example, a big photoshop filter operation might take up HUGE amounts of RAM, but only requires a very small amount of cache for the operations.

Wonder why he says it has 2 x 256mb ram but in all of the screenshots it has 1GB?

Exactly, plus he didnt run either test himself from what I can see (he certainly didnt run the Conroe test).

I can make XP MUCH faster by tweaking settings down to just what one particular program needs and can skew results immensly. This is why all comparison testing needs to be done by one small team who setup the machines the same as much as possible (obviously drivers and such need to change, but keep the OS and parameters the same otherwise).

His thinking is also "L2 cache = same as RAM but less of it and faster", which is completly incorrect.

Hellman109 writes...

I can make XP MUCH faster by tweaking settings down to just what one particular program needs and can skew results immensly

That's the beauty of SuperPi - you can't really tweak for it at all.

Its a test based on raw CPU horsepower.

Viper6 writes...

O
Viper6... sharikou.blogspot.com/20...eing-busted.html

Interesting times ahead :P

This is really interesting :):) The Conroe corrently has no performance in 64bit, well, Intel cant win this with just the 32 Bit side of things :)

LiL Macka! writes...

The Conroe corrently has no performance in 64bit

I have not seen any 64 bit tests of the Conroe anywhere.

well, Intel cant win this with just the 32 Bit side of things :)

Please acatually read what people are posting. The blogspot post is full of innacuracies, tests he didnt do, extrapolation of results from a data set of one, lies, etc.

Hellman109 writes...

Please acatually read what people are posting. The blogspot post is full of innacuracies, tests he didnt do, extrapolation of results from a data set of one, lies, etc.

Read Conroe Performance Bust thing, AMD had 10-40% more performance than Conroe in 64bit. :)

LiL Macka! writes...

Some read think to the Conroe Performance Bust thing, AMD had 10-40% more performance than Conroe in 64bit. :)

Link?

pi_1m=39.6sec(conroe 2.4G=21.25sec)

Thats a 4600+ @ 2.7Ghz Vs a 2.4 Ghz Conroe... Super Pi to 1m

www.xtremesystems.org/fo...read.php?t=95021

sharikou.blogspot.com/20...eing-busted.html

Think thats it, look for something about Conroe sucking in 64bit.

None the less a very impressive chip, a great jump from the old Pentium D and Pentium 4's nice to see Intel being praise for once aye lol. would like to see what AMD pull out of the hat, and no i am not talking abt AM2 :D. Can't wait for its offical release to the markets.

LiL Macka! writes...

Think thats it, look for something about Conroe sucking in 64bit.

The only thing about that is on a user (another blogger Ill add) saying that, saying theres no proof of it's performance. Well whop di do, 64 bit tests havent been done, hence no proof.

Here, Ill show you why my upcoming processor named the "409'er" is the best chip coming:

Quote from Hellman109:
The 409'er runs at 44 Bajillahertz with 62 Megawhallops of L1 cache and 30 Bajillion calculations per second. It can also scam you out of money faster.

OC'd Conroe does 1M SuperPi in 16s

HKEPC shows Intel's Conroe E6400 overclocked from 2.16GHz to just a little above 3.1GHz. This setup manages to calculate PI to 1 million places after the decimal point in just 16 seconds - faster than a 7.2GHz P4.

www.techpowerup.com/?11099

Viper6 writes...

OC'd Conroe does 1M SuperPi in 16s

=|

Twice as fast as my Intel at 4Ghz.

Amazing...

the future will definately be very exciting indeed :D

Viper6 writes...

OC'd Conroe does 1M SuperPi in 16s

The 6400 isnt a full feature card either... it has half the cache of the 6600/6700... amazing... Imagine those (a 6600 does it in 24 sec at stock IIRC)

samurai1 writes...

here i am thinkin of buyin a happy computer takes me two days of lookin at a screen to figure out wats good and then there conroe and am2 and vista and wateva dont think ill everget around to deciding wat i should upgrade to and wen

*note I didnt quote the start*

OK, there is never a perfect time to upgrade, ever. What is new and kick ass now is slow in 2 years guarenteed.

Right now the sweet spot is 2GB RAM, 3800+ overclocked, 7900GT (If anywhere has stock). The sweet spot changes every couple of months. This is a gaming system obviously.

samurai1 writes...

have to be <blank> rocket scientist to figure out all this stuff

No you don't. ;-)

here i am thinkin of buyin a happy computer takes me two days of lookin at a screen to figure out wats good

Don't worry. Everyone has a slow learning stage at one point. :)

Took me a while to get my current system together 11 months ago.

*Note i didn't quote the "other bits" *

sorry viper i tried eg. @# but i couldnt express myself

More Conroe Performance - Aquamark3 WR

Freecableguy of Xtreme Systems has managed a world record in Aquamark3 using Intel's Conroe CPU, clocked in at 2.72GHz, ATI X1900 cards in Crossfire on a stock Intel D975XBX motherboard - all air cooled!

www.techpowerup.com/?11192

I read this a lil bit earlier. I was skeptical about the early benchmark testing conditions placed by Intel. But the evidence seems to be mounting that Conroe will in fact trounce the K8

Conroe Vs AMD X2 and FX-57 @ XtremeSystems

First 3D Mark results Our very own X.I.P FreeCableGuy finally managed to get his sample of Conroe to work. And achieved very high scores in 3D Marks using just air cooling.

www.xtremesystems.org/fo...read.php?t=97248

Conroe breaks 4Ghz

Intel 975XBX Motherboard unmodded, no bios mods. Vapochill LS Modded by Chilly1, Memory @ 3-4-3-9 @ 1:1, Vcore @ 1.565v.

www.xtremesystems.org/fo...read.php?t=99981

i dunno how much longer i can wait for these conroes to come out

only hoping they are actually on time! :P

OMFG!!!

BeeSA EVOLUTION writes...

i dunno how much longer i can wait for these conroes to come out

Lolz, i'm also waiting for the core duo 2 to come out... Been waiting for AM2 then thought why not also wait for the intel version?
If i want to go cheapo, am2 most probably will drop in price when intel comes out. can't go wrong!

VL308 writes...

meaning amd or intel?

it depends how old u are....if your a gaming noob then intel is the original darkside, but if your old enough to know better then ud hit up a P2 over a K5/K6 anyday

Klamath writes...

it depends how old u are....if your a gaming noob then intel is the original darkside, but if your old enough to know better then ud hit up a P2 over a K5/K6 anyday

Agreed there, going by the survey though were all getting older, but it would be interesting to see how long everyone has been around PC's for.

Celeron 300a anyone?

I remember when my 386DX was the bees knees... then a mate got a new 486DX and I was shattered, had my kick ass system look dull :(

But even before that I was using C64's and such aswell....

Intel was the origional darkside for sure, I remember arguing the Amiga over the 386 (think it was the 386) back in the day...

Now it's Amiga, who they? Dont they make that physics card? No, that Ageia...

Ahh I havent had an Intel chip in my home PC since my Celeron 900 (all I could afford), so in a way it will be nice to get back to the 'dark side' again, with an Intel Chipset, Always have liked the Intel Chpsets software.

Cant wait for G80/Core2Duo

Well, I for one will just sit back and watch what happens, I won't be rushing in to buy something I know little or nothing about.

I want to see reviews with real world app's not just synthetic benchmarks, that are truly independent.

This is going to be interesting, the pressure on Intel is enormous.

The noise of the crowd dies down, the lights are dimmed, the curtains are drawn,
- what will the final verdict be ?

superpi is a load of e-wang...

any p4 beats my 2.6 ghz x2 3800+ but how many of those p4's beat mine in anything else?

im talking real world benchmarks

Hellman109 writes...

I remember when my 386DX was the bees knees

I remember replacing my 286 with the mighty Pentium 66, It use to kick the 486 DX100's.

Sadly like everything, its only good until the next gen comes out. Roll on Conroe, theres going to be a lot of very superceeded hardware out there.

Hellman109 writes...

I remember when my 386DX was the bees knees... then a mate got a new 486DX and I was shattered, had my kick ass system look dull :(

i had a 286 10MHz with 1mb of RAM....kicked a$$!
But the one that i love (and i still have the chip) is the 486DX2 66MHz, passively cooled CPU! gotta love it!