Looking really nice there. One thing you have to remember is whatevers good at spi1m never really pulls it @ 3d benching, however im waiting to be proved wrong :slap:
Printable View
Looking really nice there. One thing you have to remember is whatevers good at spi1m never really pulls it @ 3d benching, however im waiting to be proved wrong :slap:
dothan proved you wrong long time ago :)Quote:
Originally Posted by Vincentvega18
more testS?
@ vincentvga18
Dothan IS very good at Spi and 3D..........
Alex
yup that's what I was getting atQuote:
Originally Posted by AgonxOC
:banana:
VID codes go from .875v to 1.5...All within the requirements of Conroe. It's not that the BIOS doesn't support the code, the BIOS doesn't support the voltage
so, yet another day passes without any new benchmarks :( . All we gets is a SuperPi, but that can be extrapolated from Merom already. I guess well probably have to wait for the weekend before anyone will get enough time on a Conroe to really get some substatial numbers others then a few synthetic tests.
Look at it this way, if it weren't for XS we'd be waiting for months. :stick:Quote:
Originally Posted by BrownTown
Good luck with the 975:
http://www.intel.com/design/mobile/datashts/309221.htm
Section 3.10...Especially the footnotes. May be SOL
EDIT: Beware datasheet is Merom, SV requirements are the same as Conroe
Holy shat! Nice work Victor. Impressive indeed! :woot:
if you really want to show off and ruin amd, you should run the N-bench benchmark (by amd)
http://www.amdboard.com/amd_n-bench3.html
im running a 2.2 dothan w/ 2x512 800FSB 2-2-2-5, and i beat an athlon x2 3800 (2.4Ghz) by 10 points in cpu score. I can only imagine what kind of damage conroe could do.
this is mine (with my old 9800pro)Quote:
Originally Posted by deetsnai
conroe will be off the charts!
Wow I just read the whole thread, too bad results are still slim. Hope you guys can get some PI going soon, nice to see the next big thing so early. :woot:
http://developer.intel.com/design/pc...D/D1060560.pdfQuote:
Originally Posted by freecableguy
you need 10 resistors & 2 capacitors to make it work! Now you just need to find out where to place them, to give the CPU the right signal to fire up.Quote:
Originally Posted by http://developer.intel.com/design/pcn/MTHRBRD/D1060560.pdf
BTW, you didn't answer my post before about the 0403 BIOS (unreleased BIOS) for the P5WD2-E Premium. (Has anyone tried that? )
For more information see:Quote:
Originally Posted by Nasgul
RealWorldTech - Core Microarchitecture
SSE4 was originally a new extension/instruction set originally inteneded for Tejas, but once the decision was made that process/arch. would not scale much higher in clock (physics limitations), there was an abrupt discontinuation of this project (recall the news :) ). Anyway, the instructions are now located in Core, but are not intended to be significant in many ways so Intel has not really marketed as something new or cool.
wow not a bad Pi run. This Conroe is a beast indeed but could you run some 3D?
As has been mentioned before I would love to see super_pi 32M.
This would force the issue and rule out the large cache helping out.
32M would not fit in cache and therefore would stress the cpu and memory bus.
Still, I feel a strong urge to invest in my "home office" system.;) ;) ;) ;) ;) ;) ;)
he already did a 32M run, but the results were going very well so he stopped. Probably nothing new gonna be posted till tommorow, and I doubt anyone with a Conroe is gonna spend the time to read the last 200 posts and see what yall what them to bench either :p:
no 32M???
Well, they (Conroe owners) only need to read the last 2 posts to "see what yall what them to bench".:p:Quote:
Originally Posted by BrownTown
now that's better :p:Quote:
Originally Posted by garkon
bios still need to be updated.Quote:
Originally Posted by BrownTown
the system performance not so stable.
sometime fast and sometime slow.
pi_32m running very slow, so I gave up the test.
some other test result may aslo be incorrect.
sisoft cpu_alu=16784
sisoft mem bandwidth=23xx
3dmark05 cputest=1996
mem bandwidth 2.3GB/s?? something IS wrong
thanks for the update...
WOW Killer 1m PI times on these chips Victor, thanx for sharing that.:slobber:
Dying to see the 3d numbers...
I think we must bench pi_2m in the future as a standard , especially for some guys like you :DQuote:
Originally Posted by k|ngp|n
Hey, it's hard enough to explain to my wife why she needs more crunching power to surf for good deals:fact:Quote:
Originally Posted by dinos22
Awesome!Quote:
Originally Posted by chrisf6969
EDIT; 275 pages ouch - let's hope there is somethng helpful
i just tell my wife I got a new computer for her....heheh....on the last changeover i forgot to grab the Mozilla favourites so i got a few :slap: heheheh....i've got a couple of dual core HP Intel PCs lying around....but i might ditch then now and just grab some conroe ownage heheh......it hit the beta circles quicker than I thought :)Quote:
Originally Posted by garkon
where did you find this info in that PDF?Quote:
Originally Posted by chrisf6969
the way these Intels are going i think we'll have to make 8M the next 1M bench heheheh :DQuote:
Originally Posted by VictorWang
I am sure Fugger & FCG went this route before they even dropped the cpu.Quote:
Originally Posted by chrisf6969
Just scanned quickly thru datasheet - and couldn't find anything where you got htat info from - what section help you to make the above conclusion?
finally someone with a working Conroe ES. :)
kudos to Victor for testing and making sure it work first before posting anything.
super fast. Cong
you should post the "slow" 32M time anyways......Quote:
Originally Posted by VictorWang
smoking fast is all I've got to say, I wonder how AMD going to compete and respond to these kind of result from conroe....hello AMD are you listening??? :fact: Now where can I get one of these bad boys!! :D :D
oh joy.....
wonder how RAM plays in here, on his setup it only says 512MB of RAM, however, when I run SuperPi 32M on my computer the system is using more than 512MB of RAM, so that could also be contributing to the slowdown. We all talking about how much slower it is to hit the memmory than the cache, but if its having to hit the swap file its really gonna take a nose dive.
The memory performance is too slow. Probably this is the reason why sandra results are slower then 2Ghz Merom tested by FUGGER.
http://fugger.netfirms.com/deargod.jpg
http://www.xtremesystems.org/forums/...1&d=1144171503
Quote:
Originally Posted by BrownTown
I'll try to re-install the OS and set everything by deafult.
Maybe pi_1m = 19.4sec will come out ag. :D
ps: I used the pi_benching function build-in Systools to run. shows 19.454sec when the system is fresh:D
well that awsome almost 1s per cycle.......... ;)
thanks for sharing :)
Try installing the Dual hotfix - it might help a BUNCH
Any chance you can disable Apic 1.4 in that bios and install the OS with the multiprocessor HAL? I would probably try win 2K on a separate partition as well with that HAL. That have solved many dualcore/dual cpu issues in the past.
small reminder :Quote:
Originally Posted by mikeguava
i posted 2 days ago : ........ >>without extensive soldering ...good luck Charlie <<
because this is exactly what an Intel guy here in Germany told me last month :
"you will have a hard time to find an existing , unmodified board that fully supports Conroe, in fact I know none. "
lets hope you will find a solution after all the hard work and costs.
For my part would prefer a little bit more communication with the community and not this kind of splendid isolation.
greets
m
small candy for the amd fan boys , who are so sceptic about the 3d performance of the new "core duo" chips.
3DMark 2006 yonah (stock air cooler) +ati crossfire stock (only slightly tweaked)
10527 3D marks
2447 CPU
http://service.futuremark.com/compare?3dm06=157796
:slobber: :hump:
http://img.techpowerup.org/060405/10...oe_pricing.jpg
how come you haven't overclock it? :D give it some juice and crank it up yo! ;)
an an unrelated note, I think the shenanigans of the alst few days kinda show us why Conroe won't be out for a few months, seems like mobos and BIOS aren't all ready and stable. Maybe thats just 'cause people at XS can't get the latest stuff, but it seems that the platform deffinitely still needs some tweaking before launch.
we have 32 M results for a first day Yonah ocQuote:
Originally Posted by garkon
19 min
expect low 18 min soon
VictorWang
Can you run 2 super pi 1Mb and post the results here? :)
I'd like to see wattage scaling numbers for Conroe. What wattage is a Conroe putting out at say, 3GHz? :confused: :confused:
wow!!! :slobber: :eek:
now I might need to put my yonah purchase on hold.....!!
Time to look into modding the P5WD2-E to reduce startup vcore.Quote:
Originally Posted by proth
Conroes running on 975X is for sure possible - early previews of Conroe were all run on the Intel BadAxe 975X board, but had a mod to supply less voltage. I have heard from several reliable sources that the latest Rev. boards can be modded to run Conroe - but still waiting to find out what they were.
I am wondering if simple manual startup vcore reduction enough to get the chip to post.
off reading datasheet
EDIT - can't open pdf :confused: :confused:
better to let victor workQuote:
Originally Posted by PoZiTroN
yonah is 2 fast for dualpi 1 m
yonah
8M dualpi (by dennis bode hwluxx Germany)
54,4 sec
Conroe is NOT a P4. P4 is NetBurst, Conroe is not. And NetBurst was slower architecture ever since it was launced. P4 has twice the pipeline steps of a Conroe, so it is not hard to predict P4 will be MUCH slower. Still, the Conroe results are very, very impressive. It makes me remember good old P-II times. :)Quote:
Originally Posted by beta67
what about this board maybe? http://www.msi.com.tw/program/produc...il.php?UID=713 .would be nice to find out how they moded the earlier p5wd2-p to run with conroe . anyone know what was done ?
Impressed! :clap:
Quote:
Originally Posted by mine
:eek: :eek: :eek:
dual pi_8m = 54.xxx sec ??
Yonah ??
Quote:
Originally Posted by VictorWang
yes victor
T2600@3,0 stock air cooled no mods
though this ES Yonah above is about one sec slower in 1m as Hicookies one
in the Yonah competition thread
@2,9 20,4 sec vs. 21 sec
Anyone wanting one of these awesome boards can get them in a store near you, or you can wait for our equally impressive i975x chipset motherboard, which is going to be released after we upgrade our platform to support the all new "Conroe" CPU chip coming out from Intel.
That is from abits press release on 3-15.
If you can find a AW8D or AW8D-max they would probably run the conroe.
None to be found in the US,maybe some can be found in China now.
mine
Cool.
you mean the one shown at IDF? It was an Intel D975XBX Bad Axe board...Quote:
Originally Posted by vanovich
I'm trying to imagine what HIPRO5 can do when Conroe will arrive in his LAB... :slobber: :slobber: :toast:
What programs / applications will benefit from SSE4? :D
Does the Merom have SSE4 ? I would assume due to it being the same as Conroe just less Cache and FSB.
Additional Comment:
Victor I saw you had some problems with OC. But can u get ur timings any lower ? Is that posible with ur current BIOS ?
Plz someone sponsor him a 1GB STICK of PC2 8000 Corsair 1000Mhz @ 5-4-4-9 timings :P hehe..
Or just give him 1GB of DDR2 coz his sitting with 256MB haha.
yesQuote:
Originally Posted by Dagalidis
but taiwanese fanboys are very fast
a.t.m yonah @3,37
spi 32 m ..... 18.04
expect 17 min soon
Quote:
Originally Posted by mikeguava
LOL. I'll be doing that today at work. Keep in mind that spec seems to be merom and I'm not sure if the dual voltage/sleep mode applies to Conroe. One of the things I picked out that was interesting is that each core can run at different frequencies and different voltages, independantly. Pretty wild for sure. The startup sequencing looks different from presler and may have something to do with. But then again FUGGER's XPRO did get initialized but maybe that was when the MCH came out of reset and hit the DDR with some default init code to get it going. Definately has something to do with VID and power up sequenceing and/or timing.
Shuttle presented a Conroe ready XPC at the CeBit, the SD37P2.Quote:
Originally Posted by Cornelious0_0
Amazing !!!
in stock less than 22 sec ...
Taiwanese fanboys are fast but Hipro5 will MOD even mousepad to break all WR... LOLQuote:
Originally Posted by mine
amazing!
Hi Everybody! :)
I sell DFI,and Gskill and downgrade asrock 939 dual-sata2 and 3000+..
so upgrade yonah or conroe in the future? or wait?
my T2300 shipped,but not have hands yet..
aopen i975x-YDG (300$)+ T2300(180$)=480$ (I think mobo is too expensive)
or
I saw conroe mobo'a(100$?)+ 2,4ghz conroe(316$)=416$
And which is faster? Anybody tested same clock?
Conroe is faster, but you can't buy it in the next months...
dual 8m pi in 54.4 secs is BS
lets see some screen shots with PiFast
yes, that is problem..Quote:
Originally Posted by d4n13l
What difference in superpi? (yonah vs conroe)
/I heard conroe and merom is same chip/
Quote:
Originally Posted by iboomalot
Okey hmm that thing has more settings to F *** around with. Not to mention How do we know whats fast if I score 20 how fast is that ? Is there a comparable table u can put up so we can check where it will lay.:fact:
ORCBEAST
I want to see a screen shot like this one of your 54.4 secs dual Pi
http://www.fileshosts.com/DFI/NF4_SL...32m_24m19s.jpg
no way you are doing a 18 min 32m and 54.4 sec 8m it doesn't scale that way.
also says test 1 was aborted in that screen shot you posted too.
what s happen today?
1 thread with 15 pages ?
300+ posts?
50000+ views??
Victor thank you for this!!!!
:woot:
look at the doubting AMD fanboys who can't stand the idea that Intel has actually produced a faster CPU... I read these remarks across several threads
"What does SuperPI mean...AMD N-Bench V3.1 or PI Fast is better, lets see how Conroe does with those apples" :mad: :mad:
"That's not real and Super PI is for Intel only not AMD" :mad: :mad:
"He didn't say what HS he was using so the tests aren't valid"
"Those test were fake and the screenie photoshop"
"wow, that's fast :eek2: I am sure AMD will answer promptly with something faster" :2cents:
Oh don't forget AMDiots using Intel avatars...thats a good one! :stick:
"Here is AMD's new pricing structure later this year :hump: (Posted in a Conroe Super PI thread)" <--Isn't that :off:
"he didn't run the test while flushing the toilet, therefore it's invalid" To be honest this is what all the complaints read to me, LOL
Mr. mine was saying about not SuperPI but PI of Systool built-in.Quote:
Originally Posted by iboomalot
http://www.xtremesystems.org/forums/...&postcount=364
Thats so impressive and so weird vcore 8O
freecable guy check your email
freecable guy check your email
VictorWang new CPU_Z version relased that already suport conroe
Download: cpuz 1.33Quote:
News and corrections:
* Memory Command Rate timing added.
* Improved support for Intel Conroe/Merom CPUs.
* Celeron M 420/430 Yonah.
* Xeon Dual core Dempsey.
* VIA C7-M CPU detection added.
* VIA PT880, PT880 Pro, K8T900 chipsets support.
* nVidia nForce 550/560/570 (MCP55) chipset support.
* VIA VT8251 southbridge support.
* Fixed HyperTransport information misreports (in dump file).
* Added K6-2 and K6-III logos.
:toast: :toast:
Mascaras, he is already using CPU-Z 1.33 ;)
And VERY nice results VW, conroe is looking very promissing :toast:
@VictorWang
To fix cpu-z displaying wrong vcore open the cpuz.ini (in your cpuz folder) and you will see
VCore=0
set that tot
VCore=1
And than the problem should be solved.:)
Quote:
Originally Posted by iboomalot
18 min 32 ???????
yonah is scratching 17 min. air cooled
yours: amd 64 FX-60@3 GHz.....24 min 19
........ intel yonah....@3 GHz.... 18 min 05
does it hurt ?
Shame on FCG and Fugger for getting our hopes up :mad:
My advice is, next time your "source" says you *might* be able to get an ES working on a current motherboard *if* you get lucky, don't ejaculate all over yourselves with excitement, and spread the word that "CONROE BENCHES INCOMING" until you actually get something working.
I feel like this was purely a publicity stunt to increase the XS readership, and it appears to have worked. Now I just feel like I have been cheated.
LOL that's better then AMD WR 32M ;)
They didn't say benchs incoming, they said they were getting conroes. And they've said many times that they don't owe anybody anything.Quote:
Originally Posted by jebo_4jc
And why do people have to replace sex with computer hardware?
Quote:
Originally Posted by El Snorro
tried...but...no use :(
everest ultimate v2.8 can show the correct vcore
Quote:
Originally Posted by thegreek
Sorry I posted wrong link before. I'll go back & edit that post. Here is the correct .pdf about the changes to support Conroe : http://developer.intel.com/design/pc...D/D1060560.pdf
Quote:
Originally Posted by NiCKE^
yes , I know
:toast: :toast: :toast:
Wow, you sure are a douchebag over something you don't deserve. Next time, get an ES chip yourself and get them working. Oh, wait, you can't even look at an ES, let alone touch or use one. :p: ES are quirky for a reason: They're Engineering Samples. The fact that users in these forums already have them in their hands is a grace. Learn some god damn patience, kid.Quote:
Originally Posted by jebo_4jc
reinstalled the OS,
all benchmark results will be updated.
huge improvement in SISoft
Any SuperPi improvements? :D
The 0403 BIOS allows lower voltage....Quote:
Originally Posted by mikeguava
:stick: FCG, FUGGER have you tried this unreleased 0403 BIOS
superpi comes more stable, 21.4sec for 5times.Quote:
Originally Posted by NiCKE^
will come pi_32m later
With Cinebench 2003?Quote:
Originally Posted by VictorWang