VictorWang, can you run Comanche 4 Demo Benchmark ?
This is an excellent benchmark for CPU and Memory.
Download : http://www.techpowerup.com/downloads/57
Thanks.
Printable View
VictorWang, can you run Comanche 4 Demo Benchmark ?
This is an excellent benchmark for CPU and Memory.
Download : http://www.techpowerup.com/downloads/57
Thanks.
awesome results looking forward to your am2 results
Agreed.Quote:
Originally Posted by arisythila
My opinion of Conroe based on what I have seen so far is that it will most likely be the fastest chip you can buy when its released. What AMD, and indeed Intel, or whoever releases 6 months further down the line is irrelevant.
If Conroe is the fastest thing around when its released, as it looks like it probably will be - even with AM2 on the horizon, then it will be the chip to have. It doesn't matter to me whether Intel manufactures it, or AMD, or anyone else.
Well now... no yonah for me then. Will play with my Opty144 meanwhile. Conroe is gonna 0wn all
How can you say it's irrelevant what they will release? We can see the next generation of Intel is superior to the current AMD generation, what we need to note is that everything is in "beta" stage also and we don't know how AM2 "preview" rigs will perform. What we can say is that this can only get better from Intels side and wait for AM2 results to do a real compare. But it's not irrelevant since you form a opinion when you see the pre-samples. And no right now Intel Conroe owns AMD since Conroe performs better, if AMD releases a better chip they will be owning. It's probably just a matter of how you see on things since we all are different we might think stuff are relevant and irrelevant. Now lets just enjoy this and stop flamewar AMD vs. Intel and see how the Conroe performs and then AM2.
subscribed;and glad your sharing this with all of us victor!
The only problem is, if AMD guys get their unheard-of-wonder just a little late, it will have to combat quad cores from Intel..
No stoppin' on this rollercoaster.
Bleed it slowly. Then a quick deathQuote:
Originally Posted by FUGGER
Hi all,
new here and subcribed.
Thank you very much for victor for sharing the info :)
Great job :clap: :clap:
Does AMD every leak anything? Maybe. Sometimes. I know AMD has just been sitting on thier ass for 2 years. no way.Quote:
Originally Posted by mzs_biteme
Actually they make sample until they get it right, there is no way to test the CPU's and add optimizations to the CPU's until you make a "engineering sample" Engineering samples have stopped for the AM2 line, and they are starting to finalize it for release.Quote:
Originally Posted by vitaminc
I had custom laser cut plexi fan holes with a pair of 120mm fans blowing on my ram , video card and if I wasn't water cooled on a heatsink.
just because he had the slow rpm fan on the zalman off doesn't mean the heatsink wasn't getting air flow.
Still like to see some benches other than SuperPi and Thanks for your input so far Victor.
BAHAHAHAHA!Quote:
Originally Posted by vitaminc
Ignorant, please leave the internet. Now.
Conroe has a modified GTK+ fsb architecture, shared 4mb cache, four-issue per clock cycle, 14 stage pipeline, 65nm. Oh! and it has 128 bit SSE registers.
Whereas K8 has a three issue clock cycle, MOESI bus architecture, 64bit SSE registers
Go here: http://www.intel.com/technology/arch...icro/index.htm
Click on the "View flash video", easier to understand for numbskulls like you.
are you sure:slapass: :p: :)Quote:
Originally Posted by arisythila
I think things are going too far in this thread to keep the ban stick untouched. Well that's me, and I'm not being here so long. But from what I've read this months, I clearly see this is not what XS wants in the forums.Quote:
Originally Posted by Copycat
funny how the :banana::banana::banana::banana: stirrers have half a dozen posts between them ........maybe IP addresses should start getting bans
what the deuce do you mean by that? what "here"? AMDs 0.65nm is slated for Q1'07.Quote:
Originally Posted by s7e9h3n
XS AMD 0.65nm isn't "here". "Here" is stores. Right?
And the 4-5 reviews i've seen of AM2 on the net, hasn't got me any exited. Not that much better performancewise than S939. Of course, it's more "future proof".
I think this has gone far enough now.
Leave the useless AMD vs Intel debates out of this topic. All what is resulted was starting a flame war. This is the Intel section anyway, so if you wanna talk about AMD stuff, you do that in the appropriate section, but not here.
The staff also notice the lack of respect towards several X.I.P.'s here lately.
The deserved their title over a longer period of time because of their huge positive contribution towards the community. They deserve the respect we all owe them, so let's give it to them instead of insulting them.
I hope this is the last time one of the staff members needs to step in in this topic.
Any more insults of flame wars will result in a 3-day ban.
I hope this is clear for everyone.
Then you should have your access blocked for bringing up AMD as well. He was right. No one's brought up AMD more than you.Quote:
Originally Posted by arisythila
Surely you guys aren't saying that's not the case, right? It's almost worth a good laugh that bad positive info is passible while good negative info is a no-no when it comes to AMD.Quote:
Originally Posted by vitaminc
AM2 is merely a new platform for AMD to build its future 65nm CPUs that's so far down the road. We are talking a full three quarters before you see any 65nm CPU in the channel.
I bet he meant GTL+ and that's not quite the case either.Quote:
Conroe has a modified GTK+ fsb architecture
The best hope to keep the playingfield somewhat level is that nVidia will keep blocking SLI performance on Intel rigs. Then that ATI's never gets better.Quote:
Originally Posted by Victor
There's really not a damned thing to argue about when it comes to Conroe, it's Faster, Cooler, uses less power and most importantly will cost less=P I hope my access is not blocked for pointing that out.
"Truth is, Conroe is using very similar marchitecture to K8 "
Bahahaha well said!Quote:
Originally Posted by Copycat
So can we have some 01's??? :P (sorry for the persistance)
If only Intel would set asside their pride and use Hypertransport in their platforms... A free performance upgrade of as much as 30%.
FSB is a hopeless outdated technology and cannot give the performance and bandwidth needed, only with outrages clock speeds which require a lot of cooling on the northbridge.
How do you then explain the "Core" micro-arch? It uses FSB, and it sure as hell doesn't need 'outrageously' high clock speeds to crush AMD 64s with hypertransport and at higher clocks.Quote:
Originally Posted by Thorry
Hypertransport isn't free, and does take some time to implement.
Edit: And i seem to remember that Intel is working on their hypertransport (CSI) for future processors. (servers first -where it's needed the most)
The FSB is comparable to the HT (well not really the HT - moreso the onboard MCH) in performance. In single cored, it's probably better, but in say a PDEE, you notice the bandwidth limitations. But naturally in dual core dual Xeon setups it'll prob fall behind.Quote:
Originally Posted by Thorry
And besides, AMD's onboard MCH is full of troubles. 2T at >16 chips per controller, Cold bug, etc
Yeah, do a search of AMD investigating using a Co-Processor LOL! They're looking at using a chip just for Hypertransport. I too wish Intel would update the FSB completely. Then move the Hub architecture to a Switch Arch while they're at it. The DMI link already shows it can be done. It is very fast, stable and Point to Point.Quote:
Originally Posted by Salvador
Conroe will get the ICH8 version.
Signed up a few days ago but only able to post now.
This is great news. I don't care how Intel have done it - architecture smarchitecture.
This CPU looks to be leaps and bounds ahead of anything else. I just hope they don't bump up the price of the CPU and mobos when they realise that almost everyone is going to want one.
By the way, great work VictorWang. Your input is thoroughly appreciated by me and the rest of the community and I look forward to any of your future reviews as well as other contributors.
I love you Victor! Keep em comin'! :woot:
Time to get that custom license plate!
EDIT: Argh, "Victor" is taken, gotta go with "Conroe" :D
tha's the k8 not amd's next cpuQuote:
BAHAHAHAHA!
Ignorant, please leave the internet. Now.
Conroe has a modified GTK+ fsb architecture, shared 4mb cache, four-issue per clock cycle, 14 stage pipeline, 65nm. Oh! and it has 128 bit SSE registers.
Whereas K8 has a three issue clock cycle, MOESI bus architecture, 64bit SSE registers
Go here: http://www.intel.com/technology/arch...icro/index.htm
Click on the "View flash video", easier to understand for numbskulls like you.
:YIPPIE:Quote:
Originally Posted by Jupiler
Fugger, I think you should consider making it so only users with 20+ or so can post in this thread, it is rather ridiculous.
stephen, victor, fugger, and fcg, keep up the good work, I am ever appreciative for any early glimpses that I can get, unfortunately I can not say the same for others. Keep it up boys!
OMFG is very fast... Conroe is tha Powa!
Conroe XE 3.33GHz is mine... soon very soon hhahaha
Quote:
Originally Posted by RAGING DRAGON
Welcome to the forums :toast:
Afaik - prices once set by Intel usually don't change. I could only see something like that happen if demand and supply wold be out of balance and retailers can make a bit extra...
EDIT: But then - didn't Yonah go up :confused:
I think its very unlikely that intels flag CPU will cost half the price of the competitions while probably outperforming it. Its too early to tell, but i think conroes price will be much higher than expected, at least the XE.
Looks like Intel Leaped Ahead ;)
Victor,
Have you been able to run the new Sciencemark 2.0 binaries Alex provided to you. I believe he built a Pentium M and a Pentium 4 version. Those would be very interesting to see. The MD and Primordia tests are coded in x87 by MS's 32-bit compiler, while hopefully Intel's 9.0 32-bit compiler builds code that uses SSE/SSE2.
Tim
victor, I have been a long time reader of extreme (and anandtech,toms hardware, oc forums, [H] etc) but have never had any thing smart to say so I remain quite but I just want to say thanks to yourself ,fugger & fcg for sharing your ESchips, talents and findings with the enthusiast community
Keep up the good work men
gallag
i'll run it 8hours later(too tied these days and I have to go to sleep)Quote:
Originally Posted by White Wizard
thank you very much for the kindly offer.
i'll post the result and email you the txt file as soon as i finish the test
What about testing pcsx2 0.9?:D
Very impressive.
Glad to see the big I finally dropped netblargh for a real evolution of the p3, some real competition will be good, drive dem prices down again :)
Thank You Victor for your efforts. They are very much appreciated! :cool:
And, yeah, that AM2 Spi looks funky, IMHO. :confused:
Strat
I think we will see some nice improvements in the Conroes performance with new mobo and better memory :toast:
He's using a Dell POS board with Conroe....Imagine it on a 975X from Abit...:slobber:
I would prefer Asus board, but to be honest anything would be better than that Dell board :DQuote:
Originally Posted by savantu
who said it was a dell board?
AM2 part of VictorWang's testing has been moved to a new thread:
AM2 X2 4600+ with DDR2 667 brief test.... All discussion about that test HAS TO BE there, please! Thank you!
Yeah, i have a feeling AM2 may be a flop, and conroe will win this round. I have high hopes for Brisbane though. (AMD K8L X2 65NM Q1 2007 3Ghz +):slobber:
BTW I also signed up just to post. By now it is obvious that that Conroe is not another flop like netburst was. Netburst at this stage is so bad it's embarrasing. In many FPU related benches, X2-3800 = 965 XE. I only hope amd's AM2 lineup will not be as bad compared to conroe (AMD highend = Intel Low end), but so far, this may actually be the case.
I guess AMD was not expecting Conroe to perform like this. Conroe will most certainly stir up the CPU market. It seems like it is a huge boost from anything today, especially when we look at the rate the speed of CPU's has increased lately (10-15% year maybe (no counting dual cores))
This is good for the consumer -> more CPU power for those who crave it. That being said, most people with the latest and even not so latest tech now have more than enough CPU power to do whatever they need to do. There isn't really a killed app that will absolutely need the latest hardware. Almost any new midrange video card and cpu (not counting celerons) can run any game and any app at acceptible speeds. Anything towards the high end is for people that can brag about the fasterst superPi score or the highest frame rate in a game at some wacky resolution that their monitor does not even support. Of course there is a lot of software that could be faster (most of which is used to do actual work, not game).
But I understand you guys too. I sometimes wish I had a faster CPU, even though I don't need one. I guess its like the people which want feraris, when the speed limit is 100.
My point is that for now, software has a lot of catching up to do. I just don't see the need for new hardware for gaming until there are actually games out that require it, but I guess it's good that some of you do, because otherwise no one will ever develop a better software if no one will be able to run it. I myself will for sure skip the conroe generation. I might buy a K8L though, not because I need to, but just because I can
Is there any news of a newer BIOS for any of the boards becoming available any time soon? I really would love to know how well the Conroe can overclock and how much improvement it will give in performance.
Quote:
Originally Posted by Chaos-Storm
high end xfire and sli ppl would disagree with you there.
played oblivion? ;)
Not yet. I hear its good. I'll have to give it a go.Quote:
Originally Posted by mouawad
But seriously, could you actually tell the difference in today's games between single and dual 7900GTX for instance?
(not that i own this card)
I've been playing Oblivion on our work demo systems, 955XE and Crossfire X1900 with a Dell 3007WFP.Quote:
Originally Posted by Chaos-Storm
With Crossfire off 1920x1200 is ok if not great, 2560x1600 is a slideshow. Crossfire on and 2560x1600 is smooth as silk.
At home I just have a single 7800GTX and it performs admirably at 1680x1050 on my 2005FPW, upgrading to a 2405FPW soon however so will have to see how that copes.
In the end someone will always produce something to utilise every last processor cycle, pixel shader and memory access you can muster, and there are people to whom that matters, me for one :)
Todays game physics are VERY demanding, but they still suck. It's not all about pure graphics anymore. The Ageia PhysX PPU makes software developers able to produce even more demanding game physics. Even a oc'ed Conroe wouldn't be near of what the Ageia PPU can do. Hardware has got to come before software, in my opinion. Cooperation between the software and hardware developers is in any case important.Quote:
Originally Posted by Chaos-Storm
Edit: Quotations
I want to be ready to get a Conroe as soon as they are available. Is there a motherboard in the near future that can run either a Cedar Mill or a Conroe?
edit: and without any special mod. needed to do so.
thanks,
Paul (and why do people spell own as "pwn"?
Abit AW8D (max)Quote:
Originally Posted by automagic
asus P5B DELUXE
are a couple.
Anything with a 965 chipset coming should run conroe out of the box.
Might be another month or two before they hit the stores thou.
Quote:
Originally Posted by automagic
Better to be safe than sorry. Just wait until near the launch of Conroe for the better motherboards that will/should have been fully tested with Conroe. It's a long time to wait yet so although it looks extremely impressive, let's just wait what AMD have planned.
The same reason people type :banana::banana::banana::banana: as pr0n. It makes them feel "1337".
Quote:
Originally Posted by Plaicd
Abit AW8D-Max can support Conroe ?
really ?? :eek: :eek: :eek:
"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".Quote:
Originally Posted by VictorWang
http://www.abit.com.tw/page/en/news/...CNO=en_0603301
They list the AW8D on their web site as a new product but none for sale that I can find in the U.S. yet.
I would not trust any es sample to run the conroe thou.
How long has it been since DFI made a hot board on the Intel side? Perhaps the ASUS may be a better option until the DFI Super Pluto i975XX comes out, which should be revision 6 or something :slapass:
I really hope we can get a nice conroe for under $400. Im sure Intel will have each price range set. They want that mid range market back.
Victor thanks again for your tests. All we need is some super hot DDR2 to go with this chip and were golden :toast:
I dont know if this has been posted on the 30 pages but I did read that some people sayed the true performance of amd blah blah blah, well AMD aparently DOES have problems acording to anandtech:
http://anandtech.com/cpuchipsets/sho...spx?i=2734&p=2Quote:
Today's update had nothing to do with Socket-AM2 or AMD's transition to DDR2, but we do have some news with regards to that as well. As you've undoubtedly heard and maybe even seen elsewhere, AMD's AM2 platform has not been performing very well in development over these past couple of months. Just before IDF, AM2 samples were still performing lower than their Socket-939 counterparts, which was beginning to worry us and AMD's parters a bit. However we're excited to report that AM2 performance has finally started exceeding that of Socket-939. The performance gains we've been hearing and seeing are generally 5% or lower at the same clock speeds using DDR2-800, but it's still very early. The main point to take home is while the final verdict is still not out, AM2 is at least starting to look like more of an upgrade and not what we saw with Intel's DDR to DDR2 transition almost 2 years ago.
Although they are solving them.
time to save my money:). can't wait for conroe.:party: :party3: :party2:
Nobody noted, but the engineering sample stepping is 1, as opposed to prescott ES steppings of 3 or 4 some years ago. Intel made a functional process on the first stepping!!! Booting windows and yielding jaw-dropping benches...
Actual production steppings will likely be 2 or 3, which means they will have tweaked out RC delay loops, optimized power layouts slightly, and shorten critical paths not to mention optimize processing parameters.
The production samples will likely come out of the gate performing slightly better than what we have seen. Perhaps 1-5%.
Which means if productions is 3 or 4 we should have a really snazzy chip :D
you forgot the cpuz cannot recognize conroe very well :pQuote:
Originally Posted by JumpingJack
test result updates ag.
use 512m*2 crucial pc6400 :D
http://vic.expreview.com/read.php?4
Nice results Victor! What speed/timings are you running the mem?
For reference, your 1M is 1.4s faster than my 661 CM @ 5.75GHz DDR2-1064 4-3-3-4 on the XP6400s. That's huge, but my 32M is 19m 48.610s @ 5678MHz DDR2-1050 4-3-3-4. Did you lower the latencies or change the divider for 32M or something? I am surprised your 32M is slower than mine if 1M is that much quicker. Did you set LargeSystemCache = 1 in the registry for 32M? It's a huge gain on 32M.
I wish you could OC that thing, even just a little :D
Well I think Asus and Abit are much better then DFI overall. Atleast they do the tests in the Lab and not on/with us the enthusiasts. As mentioned above one should stay away from DFI for Conroe for 3-4 months as they may get a new revision with much lesser Bugs :p
Victor, can you make some game benches (F.E.A.R? HalfLife2? something?) ?
Link doesn't seems to work...Quote:
Originally Posted by VictorWang
I could see a 500 mark improvement on the CPU score and just 30 on the total score. So good boost for the cpu atleast.
Victor, plz show us 2k1 scores with that
3dmark 2006 with 2x512 ? :slobber:
Yes, please do this :). I don't want to pester you because you don't have to do anything for us, but I would have been happy with seeing 01 and nothing else, haha. It takes less time to do than most of the other benches and isn't a very large download, so if you could be so kind....Quote:
Originally Posted by \Karting_freak
Also, some archiving benchmark would be great. WinRAR 3.60 and 7zip are both multithreaded and both have benchmarking tools included.
ya, I did the OS optimized.Quote:
Originally Posted by boostedevo
but... I cannot adjust any timmings of my ram.
so, it must be run at 4-4-4-12 ddr2 667 :(
your xp6400 is soooo good that can run at 1064 4-3-3-4 , awesome:eek:
I have to download 3Dmark2001 and will test it. just let me have some rest, I'm toooooo tied these days. :(
i don't play any games, so, plz don't ask me for the gaming bench.
winrar .... i don't know how to test.
With WinRAR or 7zip you don't have to do something special, just use the benchmark option (included in program) and report the scores after 2 or 3 minutes of use.
Download Winrar 3.60 (beta I think), install it, start the program and press ALT+B. It would be interesting to know both single treaded and multithreaded scores (you have an option to disable the multithreading support in WinRAR benchmark)
@VW: Nice results sir!
It seems that everyday, the Conroe just keeps getting better and better :)
Quote:
Originally Posted by sergiu
I use winrar3.51 to test.
about 7minutes.
shows: 585kb/s
how's that ?
VW,
Should be w a y higher than 585kB/s. :(
For comparison, my singlecore Opty@2.8GHz does 619kB/s.
maybe something about HDD or Ram ?Quote:
Originally Posted by largon
my T2600 @ 3.1G with DDR2 800 4-3-2-8 1G*2 = 744kb/s
ROFL, that's the dumbest conclusion I've seen anyone come to.Quote:
Originally Posted by mzs_biteme
VW,
I have a sluggish 200GB Samsung SATA with 8MB (SP2004C), it's not the fastest HDD out there.
Different versions of WinRAR would explain but mine is also version 3.51.
why would you say that........and do you know better?.........Quote:
Originally Posted by LiQU!D
ram timings....Quote:
Originally Posted by largon
conroe's performance must be higher than Yonah.
Yonah @ 238*13, ram DDR79x 4-3-2-8 , scores=744.
so... Conroe 2.4G maybe about 700 equals to 2.8G Yonah
yeah conroe = merom......so should beat yonahQuote:
Originally Posted by VictorWang
Please try Winrar 3.60beta. Version 3.51 isn't multitreaded. Some results for comparation:
P4 2.4c@
2Ghz 1Gb ram@414 => 326KB/s
3.11Ghz 1Gb ram@518 => 413KB/s
Winrar 3.6beta
P4 2.4c@
2Ghz 1Gb ram@414 => 471KB/s (387 HT OFF)
3.11Ghz 1Gb ram@518 => 619KB/s (494 HT OFF)
It's interesting that performance does scale on multithreading (around 20%) but not on speed (at least on my CPU). Or maybe higher ram timings are the problem.
with the cpu-z verification, there's no denying it..
conroe is simply going to own..
Quote:
Originally Posted by boostedevo
Its due to the large Cache on the smaller runs like 1M
Don't blow up the problems.:stick: DFI has alot good products.Quote:
Originally Posted by harshal
Quote:
Originally Posted by sergiu
T2600 @ 236*13=3069mhz
winrar3.61
multi=1390
single=843
-------------------
Conroe E6600 @ 266*9 =2400mhz
winrar3.61
multi=1120
single=675
Hoho that's really nice Vic! :D
Can't wait to see what 3DMark01SE looks like. These Conroes are gonna be hard to beat. Thanks again for all the info Victor.
Nice results! By looking at scores, it seems that Conroe will be twice as fast as my Nordwood rig in single thread programs (and usually a Nordwood CPU is faster than a Prescott one). Also it seems that WinRAR test is ram timings dependent. I tested my sistem at the same speed with worse and good timings and I noticed 8-10% improvements. Probably a Conroe with ddr2 800 and good timings will do much better in this test.
Quote:
Originally Posted by VictorWang
very nice results vs mine
4400+ @ 256.5*10=2565mhz
winrar3.61
multi=1140
single=676
-------------------
Conroe E6600 @ 266*9 =2400mhz
winrar3.61
multi=1120
single=675
so far in SuperPi Conroe is kicking tail which Intel shines at
Cinebench95 Conroe is doing well matching my 4400+ at 2.68 ghz \
Winrar is a bit closer with the 4400+ needing 2.56 ghz
granted the above is done at crap timings 3-4-4 will revise once my OCZ-EB arrives on thursday. Need to RMA my redline too :(
Either way if Conroe's 2.4 ghz can OC to 3.0ghz on water for 316.00 , could be very tempting to change.
The results may raise a little when programs will know Conroe optimizations. Remember that there is no compiler (as far as I know) supporting Conroe while almost all compilers know AMD 64 or P4 optimizations (not only 64bits code).
Thank you very much--makes the future look pretty bright!
Vic could u try to overclock the Conroe ?
more than 3 Ghz would be :slobber:
If you read the whole thread you will know why it isn't clocked ;)Quote:
Originally Posted by Xptweaker
To make it easier for you I can say there's no options in bios and there's no software to raise fsb.
Dell motherboard hahaha. nice one.
I believe this mobo will be highly interesting. It might support Conroe:
http://www.xbitlabs.com/articles/edi...t2006-3_5.html
http://www.theinquirer.net/?article=30298
Nice finding, hope Vic can get his dirty hands on it now ! :P
Two 24-pin connectors :eek:Quote:
Originally Posted by integral
:p: im asking myself . why didnt they send one to fugger or someone here .than we can confirm for sure if it sapports conroe . and its oc capeability. then we can really see what conroe is when oc .:slapass: