Core 2 Duo is slow as hell
I was at Frys today and saw a Core 2 Duo 6400 PC. So I played with it, such as launching Internet Explorer and playing some video. The crap is definitely very slow in responding to my clicks. When I launch IE on the X2 4200+ PC next to the Core 2, it's up instantly. On the Core 2 Duo, it takes about 2 seconds. Viewing video is the same. On X2 4200+, response time is near zero. On Core 2 Duo, it's noticeable.
This brings up a question: response time and throughput. With AMD64's Direct Architecture, low latency leads to low response time. With Core 2 Duo's FSB+cache architecture, latency is large. I guess when I click the IE button, the Core 2 Duo has to do a lot of pre-fetching and caching on the FSB bottleneck, resulting in a non-responsive machine. AMD64 does not have this problem, and feel snappier.
Benchmarks mostly measure throughput, not response time.
When you play a game, it doesn't matter how large the frame rate is, as long as the rate is enough for smooth viewing. What matters in a game is the response time between you pulling the trigger and the firing of the shot. A Core 2 Duo may generate 10% higher frame rate, but it may lag by 0.5 second in responsiveness. If that is the case, the dude with Core 2 Duo will be the one that is fragged.
I just downloaded Windows Vista for AMD64 RC2 (64 bit) and installed it on my Athlon 64 X2 3800+ (which I paid $159 for both the CPU and MB with ATI IGP). I am typing this message on this machine right now. I have to admit that Vista looks great. The nice thing is Vista automatically load Cool& Quiet and the PC is running at half the speed. The CPU usage was about 10%. But you definitely need more than 1GB RAM.
Installation was quite fast, except at the beginning it seemed to be stuck without any indication of progress. One nice thing about Athlon 64 is you get native true 64 bit power, unlike those emulated 64 bit technologies.
posted by Sharikou, Ph. D. @ 10/07/2006 01:41:00 PM
89 Comments:
Ohh please, the E6300 I have is a whole lot more responsive than my 4400. Don't spread FUD, oh wait your sharikou, what was I thinking.... :)
What a load of crap. Going from an x2 4400 to e6600 is a major improvement. Maybe you're just slow at blinking, or like writing rubbish. IE loads instantly on both. Anyone with a brain would know there's other factors involved including HDD, RAM, Pagefile.
Nice title to your post...
If you had done a cold restart on both systems then maybe I would believe you.
If someone came to the AMD computer before you, looked at IE and then closed it of course it will open faster.
I say this having done the samething as you, but it was CompUSA and it was an E6300, I found it very responsive, I did not compare it to any AMD machines though :)
Is this your first foray into Vista. If it is, your soon to be disappointed. Once you get past the looks, there's not much there and a lot there that shouldn't be.
UAC sucks hard, I'm hoping they'll fix it and get some font consistency when they go RTM. Not holding my breath, I've been testing since the 4*** builds and have only gotten more disappointed. I probably won't upgrade till SP1, unless I can run Leopard on my x86 box like I can Tiger.
Load up IE once and close it. Next try to open it once more and notice the speed difference.
Of cource you will probably say that C2D will load it all to its massive L2 cahce and thus makes it fast ...
Is this your first foray into Vista. If it is, your soon to be disappointed.
I know. For us, the simpler the better. All these whistles and bells are mostly useless. But there are some improvements. For instance, I was at the control panel trying to setup remote desktop, don't know where it is, I typed "remote" in the search, and the link to setup remote desktop appeared...
On the Core 2 Duo performance, I did try a few times, and I made sure nothing heavy was running on that PC.
Shark, missed you, IV Board needs a good poster like yourself. Don't worry about the negative stuff, you don't deserve that kind of stuff.
Cheers,
plantlife
"I know. For us, the simpler the better. All these whistles and bells are mostly useless. But there are some improvements. For instance, I was at the control panel trying to setup remote desktop, don't know where it is, I typed "remote" in the search, and the link to setup remote desktop appeared..."
Your right there are some improvements, navigating through the crumb bar, new folder views/colors, etc. But the new start menu is a step back and like I said the mismatching of font is plain stupid, there's no excuse for it.
Well windows was always made more for AMD's and oviously perform better in 64bit mode. This hurts my E6400. Its not slow. Its just a little bit slower then my 3800X2 at the same speed.
This is the fault of vista and 64bit. Not the E6300 you used. Well the instructions are partually to blame because Intel never did 64bit very well. Conroe rules the 32bit systems. AMD the 64bit systems!!!
Yo Pretender..
This is the same kind of deep research that earned you that PhD you so proudly add to all your signatures.
I find in interesting that those who know the least need to prove themselves with fancy titles and loud claims. Sharikou you prove it again and again..
Prove me wrong and actual post something that makes sense.
Yup INTEL going bankrupt, Yup INTEL CPUs cause fires, Yup all the world is wrong about Core2 CPUS they are dreadfully slow. Its a horrible and evil plot by Damm Hateful Pat G.
You are a joke and a disgrace to anyone who has a real PhD.
Oh, please.... Sharikou. Dont do this. Everybody knows the difference between Conroe and AMD64 X2. If you continue doing this negative advertisement, you risking to lose your credibility among persons they believe you. Conroe is nice fo now, is going to be shiny for a while, but everybody knows, especially in this computer business, nothing last forever....today uphill, tommorow downhill....
"Ohh please, the E6300 I have is a whole lot more responsive than my 4400. Don't spread FUD"
Oh yeah? And YOU are not spreading FUD by saying that above?
Sharikou was simply stating his experience. It was not scientific, and it is indeed misleading. But at least he is better than some people who fake about his experience at all.
The E6300 more responsive than the x2 4400 are you kidding me? The E6300 takes the FX62 out back and beats it like a stepchild. Core 2 is in a league of its own right now. Core 2 rules this year, but next year it will be K8L turn to take Core 2 duo out back and beat it down....can't wait.
"Ohh please, the E6300 I have is a whole lot more responsive than my 4400. Don't spread FUD"
Oh yeah? And YOU are not spreading FUD by saying that above?
Sharikou was simply stating his experience. It was not scientific, and it is indeed misleading. But at least he is better than some people who fake about his experience at all.
I was stating my experience as well. I personally own two rigs, one is an hp pavillion x2 4400 and the other is a e6300 that I built at home. Both are great machines but the e6300 is a whole lot more responsive than the hp rig. Now I understand that there is a whole lot of different factors involved here so chances are that since I just installed my e6300 with a fresh xp install that it will probably be faster. Lets be honest here, doesn't your computer always run faster after you format your drive and install a fresh copy of windows? So I guess if I'm spreding FUD here than so is the blogger, no?
1. No one should believe anything that you say. You would be the last person to be in Fry's ;) PS, I'd like a photo of the oh so mythical Doctor Sharikou.
2. Native/true? EM64T is 'Extended Memory' not emulated. It's not anymore different than AMD64, just branding.
You know what is funny? Intel stands up there, runs an app, tells you the exact configuration they used, and you call it gurrilla benchmarketing. But then you claim that you walked over to fries and you thought that C2D was not responsive, and hence it is slow as hell! What a boat load of crap!
BTW, I am typing this message on my C2D E6700 (non OCed), while I am encoding 2 hr length movie in the background (64% encoding already complete in 40 min), and yet my IE is pretty responsive.
What Sharikou says is utter lies. I know umpteen guys who hv moved to Core 2 from Athlons. They are all are very impressed. Not just with general usage but games as well.
Relying on faulty logic is one thing. But now Sharikou is relying on complete lying to spread his message.
I honestly don't believe Shakira went to Fry's ;) And I don't believe he has any credibility or any followers besides MadModMike. It's just fun to point out his BS:D
I am not surprised that Sharikou encountered a slow Core 2 Duo. It's purely circumstantial.
Nasty combination against Core 2 Duo and Pentium CPU is:
_ Application is bigger than the tiny cache.
_ Application's algorithm cannot utilize pre-fetch to high latency.
Such combination will expose and exasperate the weakness of not having on-chip memory controller.
While Core 2 Duo is fast alright, but it has the above Achilles' heel. Not only that I despite the brute-force approach from Intel, I cannot depend on a product with Achilles's heel.
-Longan-
idiots e6300 and e6400 suck
only at e6500 and e6600 does conroe become worth getting and thats a fairly small segment of the market
Sorry Sharikou, it was an old socket 754, Athlon 64 2800+ machine which was on Sale. When I saw you, I switched the configuration stickers to cheer you till K8L intro.
Good luck!
Not a very scientific test but I guess it does not have to be. Another non-scientific test would be to ask how many people who moved from AMD64 to Core though core slow as help complained and then moved back.
Would the answer be approximately zero?
This anti Core sentiment is somewhat like King Canute trying to hold back the tide ...
Core2 has significantly (or at least "noticably") lower latency and superior 64-bit performance.
These are facts proven by 1 million instances, while your opinion is just FUD spread by a retarded hypocrit that blames others for his own failure.
Benchmarking at Fry's... Is this what has become of your ambitious attempt to have a benchmarking web site. Have you really been reduced to writing about your "feelings" while loitering about at Fry's? How sad. Do you take handouts err, donations on your website? I want to give you some money so you can get something to eat Sharikou.
If you continue doing this negative advertisement, you risking to lose your credibility among persons they believe you.
Comments like these always make me smile... One thing to know about Sharikou is that he doesn't have any credibility left to lose so he can say whatever he likes with no consequences. It must feel very liberating have such low credibility.
I don't understand how you could attribute a noticeable lag in responsiveness to the latency of the microprocessor. The largest latency for the processor would be about 50 nanoseconds. You are talking about a lag 10 Million times greater. The lag could have been caused by lost resources or even by adware. It could also have been caused by too many resident applications or toolbars. I just don't understand how you can relate two things of such vastly different magnitudes. Maybe you believe in the "butterfly effect".
" S said...
What Sharikou says is utter lies. I know umpteen guys who hv moved to Core 2 from Athlons. They are all are very impressed. Not just with general usage but games as well.
Relying on faulty logic is one thing. But now Sharikou is relying on complete lying to spread his message.
12:25 AM, October 08, 2006 "
I've seen quite a few ones on tomshardware.com bitching about how hot the core2duo can be...
some are overing 60 degrees celsius..under non overclocked stuff, others show random crashes.
so dont say "everyone"
Anonymous said...
I honestly don't believe Shakira went to Fry's ;) And I don't believe he has any credibility or any followers besides MadModMike. It's just fun to point out his BS:D
12:37 AM, October 08, 2006
Im still wondering, if you guys hate sharikou that much, then why the hell are you on his blog?
it doesnt make sense!
Some monkey said
"If you continue doing this negative advertisement, you risking to lose your credibility among persons they believe you."
Are you that ignorant? Sharikou the Phd pretender makes it his business to maximize false advertising and ignoring facts, and obvious trends. Kind of like his Phd Pretender title.
The more you spew the bigger the joker you prove yourself to be.
You every figure that out yet? Or do you realy believe that PhD is some sort of title that gives you right to think illogical and type like a moron?
I have not seen one logical post from you in the few months I've been entertained by this site. You are like a bad TV show.. worthy of somes side intertinament
"A Core 2 Duo may generate 10% higher frame rate, but it may lag by 0.5 second in responsiveness. If that is the case, the dude with Core 2 Duo will be the one that is fragged."
Umm, you do know that input lag and framerate are directly connected to each one and other? The higher the framerate the better the responsiveness. Thus higher framerate translates directly to better responsiveness.
With desktop apps CPU is the last thing to slow things down. The first thing is HDD. If you need to access that you cannot make a difference between quadcore and original Pentium or K6.
"idiots e6300 and e6400 suck"
Actually they offer the best bang for buck of all the C2D's. e6600 is not worth all that extra money since doubling the L2 cache will not help all that much.
If you intend to OC these Allendales are the best dualcore CPU's assuming your motherboard can tolerate 450MHz+ FSB. It is rather simple to get e6300>3.1GHz and e6400 >3.5GHz. All you need is a decent motherboard :)
"I've seen quite a few ones on tomshardware.com bitching about how hot the core2duo can be...
some are overing 60 degrees celsius."
Interesting. My e6300 runs 24/7 under full load for weeks in a row and so far highest I've seen has been 40C and that was when I had accidentally blocked the only place wher air could get out from my case. Also that CPU use OC'd >70%, running rougly @3.15GHz on air. ATM it is running S@H on both cores and CPU is 27C, room is around 21C and CPU fan is on minimum speed.
"if you guys hate sharikou that much, then why the hell are you on his blog?"
I'm here just for the laughs. Seeing someone miss the 101 on hardware and software confirms I'm not the stupidest person alive :)
Benchmarking at Fry's... Is this what has become of your ambitious attempt to have a benchmarking web site
I was planning to do a web site. But after both IBM and DELL voted yes to AMD64, there was no more need. DELL says AMD64 is the future, IBM says similar things, SUN says no to Intel. You can actually get a 8P 16 core Opteron machine from SUN for free trial. Why bother with a test of my own.
As for my experience with the Core 2 Duo, it was first hand experience. I was really expecting the Core 2 Duo to be fast. But the facts don't lie. User experience is different from benchmark numbers. My user experience with Core 2 Duo is that it sucks. It's simply non-responsive and unsuitable for user interface interaction. Maybe it's good for encoding video where reponse time doesn't matter, but as far as responsivness is concerned, it's crap.
I challenge the folks with Core 2 Duo to perform similar tests: click on applications and see how quick it starts...I ensure you that you will see a noticeable lag.
Nasty combination against Core 2 Duo and Pentium CPU is:
_ Application is bigger than the tiny cache.
_ Application's algorithm cannot utilize pre-fetch to high latency.
Core 2 Duo is good for repeated tasks, such as looping through SuperPi, Core 2 Dup can look ahead and load stuff into memory. But in case of user interaction, nothing can be predicted, that is where Core 2 Duo sucks. It has no way to know which button I am going to click, and when I click it, Core 2 Duo takes a long time to respond. That's expected behaviour.
My e6400 just was not as reliable as my am2 4600 and I just took my last conroe e6600 out of service cause it cant run basic programs correctly and replaced it with a am2 5000.
I read the US Govt wont allow conroes to be purchased either.
Maybe the only software that core2s run are the benchmarks, maybe it was never tested with actual productivity software.
I read Core2 conroes are just pentium 3 chips reworked.
"I challenge the folks with Core 2 Duo to perform similar tests: click on applications and see how quick it starts...I ensure you that you will see a noticeable lag."
This was based off your 'case study' of one machine, one day at Fry's?
How stupid to you think your readers are?
Umm, you do know that input lag and framerate are directly connected to each one and other? The higher the framerate the better the responsiveness. Thus higher framerate translates directly to better responsiveness.
Not true. In a benchmark, you are running a pre-determined sequence. In a real game, the user's input causes interrupts in the operating system, then the program is called on to respond to that. It's completely random to the program. When you pull the trigger to fire a shot, that's an asynchronous event.
The difference is between a batch job and real time job. If you understand.
I don't think it's a coincidence that Sharikou of all people has something to complain about Core 2;)
"Im still wondering, if you guys hate sharikou that much, then why the hell are you on his blog?"
Perhaps we would like to see the world be rid of BS and let others see the light:) I don't hate Sharikou, he makes the rest of us look good;)
Sharikou WINS! Intel to go BK in 2007. In other news a man reported that he saw pigs fly :) Oh and by the way I'm very happy with my e6700 rig. Very responsive and runs cool at 33c idle and 45C during load, OC to 3.5Gh z on air. Oh and by the way I did see you at frys buying a new Core 2 duo machine....good for you sharikou, about time you wisened up boy.
I think Sharikou got a point.
My friends are complains about core duo in terms of loading time in games,
They say core duo need a lots of time compeer to A64.
A few of them say to me, core duo is fast but in windows environment the cpu act like P4.
After herring this kind of stuff, better stick with AMD.
Sharikou complaining about an intel processor nah....the man is the epitomy of fair and balance. Just like FOX news channel. This guy is the Ann Coulter of the geeks.
Anonymous said...
The E6300 more responsive than the x2 4400 are you kidding me? The E6300 takes the FX62 out back and beats it like a stepchild. Core 2 is in a league of its own right now. Core 2 rules this year, but next year it will be K8L turn to take Core 2 duo out back and beat it down....can't wait.
STFU Thats a bit overdoing it in the missleading department don't you think. Your a real big intel fan arn't you? The most load of crap I ever heard. Be more logical and stop it with the BSYFH. The speed difference is roughly only 250mhz in 32-bit and non existent in 64-bit.
Anonymous said...
I honestly don't believe Shakira went to Fry's ;) And I don't believe he has any credibility or any followers besides MadModMike. It's just fun to point out his BS:D
Don't hold your breath DW!!!
Anonymous said...
Some monkey said
"If you continue doing this negative advertisement, you risking to lose your credibility among persons they believe you."
Are you that ignorant? Sharikou the Phd pretender makes it his business to maximize false advertising and ignoring facts, and obvious trends. Kind of like his Phd Pretender title.
The more you spew the bigger the joker you prove yourself to be.
You every figure that out yet? Or do you realy believe that PhD is some sort of title that gives you right to think illogical and type like a moron?
I have not seen one logical post from you in the few months I've been entertained by this site. You are like a bad TV show.. worthy of somes side intertinament
STFUB
It's a pity that Intelers have nothing to refute my observations, only their party cliche...
AMD's Direct Connect Architecture is the future. Intel will be lucky if they succeed in copying that by 2008.
"It's a pity that Intelers have nothing to refute my observations, only their party cliche..."
Total BS claims are unrefuteable.
I went to CompUSA today and compared an Athlon X2 with a C2D. Boy the Athlon was SLOW...
Therefore, Athlons are crap.
Now doesn't that sound totally stupid?
Sharikou said...
"I guess when I click the IE button, the Core 2 Duo has to do a lot of pre-fetching and caching on the FSB bottleneck, resulting in a non-responsive machine."
What kind of stress does IE put on the FSB, thats a sad example.
"When you play a game, it doesn't matter how large the frame rate is, as long as the rate is enough for smooth viewing."
So those benchmarks in the past with K8 beating Netburst are irrelevant when it comes to gaming?
I can't say I have ever noticed a lag when using my computer for gaming, if there is a lag in gaming its from being online.
"If that is the case, the dude with Core 2 Duo will be the one that is fragged."
Sharikou do you play any online games, please say you play BF2?
If so what is your nickname, and where do you play?
PS: About Vista, I did the same, but I can not get my X-Fi card working, that OS is no where near ready IMHO.
LOL..
Sharikou.. we don't need to send you squat.. The reviews are in from many different independent reviews of the Core2 vs AMDs best. All of these same organziations used to flat out tell about how AMD FX used to kick INTEL's butt. Times have changed and they all report on how much better the Core2 is compared to the AMD FX.
No refute is necessary as it is all aroudn you. You are just too blind and stupid to see it.
Benchmarks? We don't need no stinking benchmarks. Sharikou's unbiased and unwaivering knowledge is second to none. Heck I'm waiting for his indepth coverage of HDTV plasma screens at Best Buy before I go out and buy one.
The future Sharikou? Please oh great wise sharifraud tell us more about the future? Will the Bears win the Super Bowl? Who will win the world series? Will I finally get a raise this year? Please great future teller of the future, tell us more.
This has to be the lamest attempt at a blog I have ever seen. Blogs are for politics not for geeks. You nerds with e-penis syndrome need to get out of your parent's basement and get some fresh air. And no Circuit City and Frey's does not count.
WOW! Sharikou thanks for recommening the amd machine at Frys. I went in the other day and found a sempron $299 PC that opened IE incredibly fast. It was almost instant. From now on I will tell my freind to use the sharikou IE test to make sure that they are buying the best computer they can get.
" Scientia from AMDZone said...
I don't understand how you could attribute a noticeable lag in responsiveness to the latency of the microprocessor. The largest latency for the processor would be about 50 nanoseconds. You are talking about a lag 10 Million times greater. The lag could have been caused by lost resources or even by adware. It could also have been caused by too many resident applications or toolbars. I just don't understand how you can relate two things of such vastly different magnitudes. Maybe you believe in the "butterfly effect".
8:59 AM, October 08, 2006 "
Intelers don't need to, your own team mate says it right there.
I don't understand how you could attribute a noticeable lag in responsiveness to the latency of the microprocessor. The largest latency for the processor would be about 50 nanoseconds. You are talking about a lag 10 Million times greater.
Not true. First, the latency on Core 2 Duo is much higher than 50 nano second. Secondly, when mouse click trigger many many memory accesses, in the OS and the software itself. The latency can add up.
I've found the same to be true most of the time when I'm at CompUSA or Best Buy (I used to be a computer salesman there.) It has nothing to do with the processor, but just the way the companies load a million crap programs by default. For some reason the manufacturers like to load more crap on the intel machines it seems. On a clean Windows install you wouldn't notice these same things, but the Intel machines are certainly slow as hell from OEMs compared to their A64 counterparts.
enumae said...
"Sharikou do you play any online games, please say you play BF2?
If so what is your nickname, and where do you play?"
So what is your nickname in BF2? Were do you play Enumae? lol.
X-fire?
Clan?
Leagues?
Ladders?
"In a real game, the user's input causes interrupts in the operating system, then the program is called on to respond to that."
Woha, where did that come from? Even in the dark ages of DOS things weren't like this. They have never been like this in real-time games.
"It's completely random to the program. When you pull the trigger to fire a shot, that's an asynchronous event."
Yes, keypresses are gathered asynchronously but not by software.
Ever noticed that when you hold down a key on your keyboard the PC speaker starts to beep eventually? That means the HW buffer is full of scancodes and to read any other keys it has to be emptied by software.
Take your average game and you'll see a function like this:
void mainloop(){
while(true){
doAi();
processInput();
doPhysics();
drawFrame();
}
}
Of cource some games do some of these things in parallel threads but it won't change the general idea.
I'm willing to bet there are no real-time games that process input more than once per frame. A frame is defined as a single logick tick (here doAI() and doPhysics()). Skipping some drawing calls don't count.
Please, if you don't know the subject don't make stuff up. You don't seem to know much about game engines but if you really want to I can explain some basic principles so you won't make an ass out of yourself next time.
LOL! Yes, C2D makes you lag 1/2 second in games!
ROFL!!!
I normally don't post much, but this is the FUNNIEST blog entry in a long long time.
:D :D :D
Anonymous said...
"So what is your nickname in BF2?"
enumae-k
"Were do you play Enumae?"
I will have to look it up, sorry.
Sharikou,
You may have had a different experience, but I benchmarked my little 6300 with a few different benchmarking tests and the results are:
1. There is nothing that AMD offers that can come close to the performance from my e6300.
2. Any questions?
Here's the scoop - for whatever reason, your 3800 may bring up the windows quicker - but a Conroe will bring up the window, perform the task, and close the window FAR quicker than an AMD machine can. Isn't that what counts?
If you look at gaming, I have Oblivion running at ultra-high graphics mode with a X1600 ATI card (nothing special) and never have any frame rate issues. In fact, I can do something like check viruses AND play Oblivion with no frame rate loss (with only 1Gb RAM and old ATA hard drives, no less.)
If you look at text editing (perhaps more your domain), a Pentium 166 will suffice. Or a 3800 - does it really matter? You can open windows as fast as you want with any machine! Go to town.
There is nothing that AMD offers that can come close to the performance from my e6300.
This is simply false. Conroe Xe 6800 is on ly 10% faster than FX62. Are you saying Con XE6800 is less than 10% faster than e6300?
void mainloop(){
while(true){
doAi();
processInput();
doPhysics();
drawFrame();
}
}
Crap. This is the most primitive way of doing things. Based on this code, the input processing will be delayed by the physics and draw functions. Modern games uses events and handlers. Drawing is just taking snapshots of the simulated world...
This is simply false. Conroe Xe 6800 is on ly 10% faster than FX62. Are you saying Con XE6800 is less than 10% faster than e6300?
Not saying that benchmark suites offer the best 'real-world' results, and I don't own a FX62 to begin with, but according to sisoftSandra my e6300 overclocked 10% (my only caveat) cleans up on the fx2-62 in processor multimedia performance benchmarks.
Integer performance:
me - 112469 it/s
fx2-62 - 52653 it/s
X2 4000+ - 37609 it/s
c2D XE X6800 - 159122 it/s
NOT a typo - performance over double that of the fx-62 with my puny e6300.
Floating point performance:
me - 60837 it/s
fx2-62 - 56980 it/s
X2 4000+ - 40700 it/s
c2d XE X6800 - 85786 it/s
Again, no typos. Even my little e6300 cleanly beats floating point performance of the fx-62.
I don't have a 3800 in my list to compare to, but I suspect it'd be around the same at the 4000+. Either way, a 6300 modestly overclocked will soundly beat a fx-62 at 1/4 of the cost.
I have screenshots if you want me to email them to you. If you want me to do more tests I will. I suspect the fx-62 will do better in raw FP performance than the e6300, but in integer calcs the e6300 should easily beat the fx-62. Actually, I'll do that and post the results soon.
I really think that the jerky performance of the conroes is due to the lack of modern memory controller.
In all memory bandwidth tests the AMD always outperform the Intels legacy off board memory controller.
Plus in testing the e6400 and e6600 against the AM2 4600 and AM2 5000 I noticed that there are major performance lags caused by the conroes system of cpu divisional power assignments.
The AM2 is much more even in its cpu task power assigns then the conroe.
Linux and the top command show that conroes have a 75%-85% burst of cpu power assign to graphics tasks the second any occur leaving very little left for the task of running the program behind the graphics actions. This becomes very pronounced when you use a VIA chipset motherboard and a conroe cpu.
The lag in any non graphics task becomes massive when running a conroe on a VIA chipset board according to the top command analysis.
So its probably conroes cpu priority responses that make it seem so jerky and lag at times.
Simply put the AMD hypertransport is a far superior performing memory system.
Okay, benchmarks for processor arithmetic are done.
Dhrystone ALU:
e6300 oc 10% - 18938 mips
fx2-62DC - 20225 mips
C2 XE x6800 - 26501 mips
X2 4000+ - 14446 mips
Whetstone iSSE3:
e6300 oc 10% - 12976 mflops
fx2-62DC - 17084 mflops
C2 XE x6800 - 18441 mflops
X2 4000+ - 12203 mflops
So the fx-62 is a bit better than an e6300 modestly overclocked in both tests (especially in floating point calcs). And the XE6800 kicks both. The 4000+ looks pretty decent on flops and mips but still relatively low - I'd like to see how the 3800s compare, but I don't think it'd be anywhere near a conroe for performance.
I'm not biased, but I'd say a conroe is pretty decent bang for the buck. You can keep opening and closing windows with a 3800 if you'd like, but I'll pick a Conroe any day.
According to HardOCP..
http://enthusiast.hardocp.com/article.html?art=MTExMCwyLCxoZW50aHVzaWFzdA==
Again, we see that even that the most common video editing programs benefit from the Intel Core architecture. Our Intel X6800 enjoys a real time savings advantage of about 25% compared to the AMD FX-62.
You can keep opening and closing windows with a 3800 if you'd like, but I'll pick a Conroe any day.
This statement is simply stupid. Do you run sisoftSandra all day or open/close windows all day? I don't care about Sandra, I do care how fast the machine responds to my mouse clicks.
"I have Oblivion running at ultra-high graphics mode with a X1600 ATI card (nothing special) and never have any frame rate issues."
Funny, people with 5x more powerful GPU's are saying they have problems with that game. What resolution are you using and what is the average and minimum FPS you see in forests?
"Based on this code, the input processing will be delayed by the physics and draw functions."
Yes, as I said input is almost* always gathered and processed only once per frame. I've never head someone making anything else in regular FPS/RTS games. Why shouldn't it?
*) With games that use dirty rectangle systems and no realtime simulations there really are no real frames and there things might work with events. Most such games are 2d turn based a'la Xcom from '93 though I'm quite sure even there input was gathered once per frame. Just that it would be gather+respond to do it asynchronously too if someone would really be that stupid to waste time on it.
"Modern games uses events and handlers."
Modern games such as ...? Where is your knowledge coming from?
"Drawing is just taking snapshots of the simulated world..."
So it is and it goes the same for physics. Can you alter the snapshot in the middle of taking it without royally screwing things up?
If you said "yes" then tell me what would happen if you render an enemy model, read input in the middle of frame and find out that the player has just fired the railgun. Should you magically eraze the enemy and replace it with blood and guts flying around?
"I really think that the jerky performance of the conroes is due to the lack of modern memory controller."
no, No, NO!
Memory latency and bandwidth can be smaller but it will never make a noticeable difference in opening a browser window. A single disk access takes roughly 2.5M CPU clock cycles on a 3GHz CPU. L2 cache miss will take around 300 cycles on external memory controller and ~120 cycles with IMC.
How can someone even imagine CPU can be a bottleneck in that kind of situation?
"Linux and the top command show that conroes have a 75%-85% burst of cpu power assign to graphics tasks the second any occur leaving very little left for the task of running the program behind the graphics actions"
Are you sure this comes from CPU and not from motherboard or badly configured software? I've seen 500MHz P3 taking over 2.3GHz Athlon XP under Linux when the latter had a tiny configuration issue. I haven't seen such weird behaviour on my E6300.
You can keep opening and closing windows with a 3800 if you'd like, but I'll pick a Conroe any day.
This statement is simply stupid. Do you run sisoftSandra all day or open/close windows all day? I don't care about Sandra, I do care how fast the machine responds to my mouse clicks.
Yeah, go hard then. Click away. -I- happen to care about shaving a few hours off video encoding, or how when I'm listening to MP3's I'd rather them not stutter when doing other tasks, or on the rare chance I get to play a game it doesn't chug out and croak.
I'll be happy to concede when it comes to app startup or clickity-click times as long as in return I get the job done significantly quicker. However, I think the human experience has tainted your view and there is only a perceived difference. Really.
In my case, I have 6000 fonts to go through when starting up gimp/photoshop (Adobe apps were bastards for me with so many fonts), now I see a dramatic improvement at startup, but I doubt you saw any difference on a Conroe for click response time than you would on an AMD machine.
I suspect it's more of a HD access time that you noticed rather than a lack of CPU power. It seems pretty childish of you to base an opinion on a brief experience at a computer store rather than do your own tests.
Why can't you just give in and say, "Intel wins this round...wait for AMD to retaliate"? Seems to me you've gone beyond the ridiculous to pick at any little thing you can to prove that Conroe is terrible.
And no, I don't run Sandra all day of course. But I do a lot of other demanding tasks that require real horsepower, and this Conroe does a fantastic job at a low price. Mouse clicks might add up for other people, but not for me. YOU were the one bitching about mouse clicks, not me. I have real concerns.
"I have Oblivion running at ultra-high graphics mode with a X1600 ATI card (nothing special) and never have any frame rate issues."
Funny, people with 5x more powerful GPU's are saying they have problems with that game. What resolution are you using and what is the average and minimum FPS you see in forests?
1024 x 768 Ultra high video quality HDR. I'm not sure the frame rates, I'll post them when I have a chance to look. Beyond my perception capability though. But no lag, no problems.
I've played a few times right through a complete system antivirus scan though without noticing (AVG, not Norton or something horrible though. Again, only 1Gb of RAM, Radeon x1600 Pro (256Mb, PCIE-16). Nothing fancy by any means. The whole system, including the video card, 1 gb noname DDR2 RAM, a 350w PSU, an e6300, and an asus mb cost me around $600 CAD (~$550 USD). I popped in old HD's and used an old chassis.
Granted, you could get a 3800+ complete system w/LCD monitor for that price, but the performance wouldn't be as good as this.
Pretty amazing considering
"1024 x 768 Ultra high video quality HDR"
So around 15 FPS is beyond your perception capability?
http://www.anandtech.com/video/showdoc.aspx?i=2746&p=6
Though you should be getting far less since they are running at medium and you on ultra high.
"1024 x 768 Ultra high video quality HDR"
So around 15 FPS is beyond your perception capability?
http://www.anandtech.com/video/showdoc.aspx?i=2746&p=6
Though you should be getting far less since they are running at medium and you on ultra high.
Here's the results using fraps.
In Bruma:
2006-10-09 16:04:18 - Oblivion
Frames: 999 - Time: 29165ms - Avg: 34.253 - Min: 28 - Max: 45
Outdoors (outside Bruma):
2006-10-09 16:05:14 - Oblivion
Frames: 127 - Time: 4650ms - Avg: 27.311 - Min: 26 - Max: 29
Fighting outside Bruma:
2006-10-09 16:05:25 - Oblivion
Frames: 118 - Time: 6448ms - Avg: 18.300 - Min: 17 - Max: 21
All effects on full running in Ultra-high mode, HDR, 1024x768. Lowest I could hit is 17fps while fighting with spells and so forth.
So yeah, pretty much what they got although they were running in medium quality. I can't perceive much of a difference between 30 and 20 fps because I'm not a hardcore gamer. Not bad for a budget CPU, budget GPU, and half the RAM.
I wouldn't turn down the options to achieve any more than 30fps, it's not worth it to me. And I CERTAINLY wouldn't pay the big bucks for a FX-60 like their test machine. But with a better GPU I'm sure this 6300 would achieve the same as a FX. Speculation though.
"no, No, NO!
Memory latency and bandwidth can be smaller but it will never make a noticeable difference in opening a browser window. ... L2 cache miss will take around 300 cycles on external memory controller and ~120 cycles with IMC."
According to your numbers, it takes roughly 20ms to fill up 2MB L2 cache with 32B block size on a 800MHz FSB ((2M/32)*300/800M), suppose there is no thrashing at all (that's why the large cache and its high associativity are so important to Conroe).
Maybe you won't notice it during context swithces, or maybe you don't do context switch that much at all (like most benchmark softwares). I don't know.
Hey, I downloaded Vista version RC3 here
" Anonymous said...
Some monkey said
"If you continue doing this negative advertisement, you risking to lose your credibility among persons they believe you."
Are you that ignorant? Sharikou the Phd pretender makes it his business to maximize false advertising and ignoring facts, and obvious trends. Kind of like his Phd Pretender title.
The more you spew the bigger the joker you prove yourself to be.
You every figure that out yet? Or do you realy believe that PhD is some sort of title that gives you right to think illogical and type like a moron?
I have not seen one logical post from you in the few months I've been entertained by this site. You are like a bad TV show.. worthy of somes side intertinament
10:45 AM, October 08, 2006 "
Dear Anonymous!
Acctually it was my first post, so you just proved nothing else that YOU are that person youre talking about. This place starts to be very offtopic because of people being busy flaming. But I have a message for you. :" bazd meg a qva anyád, te köcsög", just for monkeying.
For you Sharikou:
Keep going..:) the Force be with you!
Bogomips, in linux bogomips are related to the speed at which you processor can process. The higher the value the better it is for system speed.
Am2-5000+ yields a 5207.28 bogomips score
E6600 yields a 4791.87 score.
Higher is better performance and as you can see by the figures above that the AMD AM2-5000+ outperforms the E6600 conroe core 2 in the bogomips benchmark.
You can duplicate this test on any linux system in the world.
This is just one more benchmark where conroes do not beat AM2s.
You fanboys keep raving about the conroe while overlooking its many short comings like, cant run basic code without errors, weak on bogomips benchmark, poor memory bandwidth.
Remember, conroe is nothing but 2 supercharged pentium 3s which basis is quite old technology.
No matter how fast you push a model t, in the end its still just a fast moving antique.
Ahh yes the bogomips benchmark...that along with the sharikou IE click benchmark will be the standard for benchmarking processors from no on.
"Bogomips, in linux bogomips are related to the speed at which you processor can process. The higher the value the better it is for system speed.
Am2-5000+ yields a 5207.28 bogomips score
E6600 yields a 4791.87 score.
Higher is better performance and as you can see by the figures above that the AMD AM2-5000+ outperforms the E6600 conroe core 2 in the bogomips benchmark.
You can duplicate this test on any linux system in the world.
This is just one more benchmark where conroes do not beat AM2s.
You fanboys keep raving about the conroe while overlooking its many short comings like, cant run basic code without errors, weak on bogomips benchmark, poor memory bandwidth.
Remember, conroe is nothing but 2 supercharged pentium 3s which basis is quite old technology.
No matter how fast you push a model t, in the end its still just a fast moving antique."
Yes the K8 is the performance champion based on the bogomips all those professional review sites are wrong when they say the conroes beat the k8 by an average of 20% in all benchmarks and programs. YOU SIR ARE AN IMBECILE spew your garbage elsewhere.!!!!
"According to your numbers, it takes roughly 20ms to fill up 2MB L2 cache with 32B block size on a 800MHz FSB ((2M/32)*300/800M), suppose there is no thrashing at all (that's why the large cache and its high associativity are so important to Conroe)."
Jeez, you don't know anything about how caches work, do you? Have you ever heard of burst speed or block reading? If stuff needs to get to cache and it isn't already there there is a single cache miss since you have to send some commands to the memory controller for it to start sending data. Once the commands are there the buses are pretty much maxed out and run very close to 100% efficiency filling the 2M cache in a couple thousand ns, maybe faster. Also there are no 800MHz C2D's at the moment as far as I know.
"Bogomips, in linux bogomips are related to the speed at which you processor can process. The higher the value the better it is for system speed.
Am2-5000+ yields a 5207.28 bogomips score
E6600 yields a 4791.87 score.
Higher is better performance and as you can see by the figures above that the AMD AM2-5000+ outperforms the E6600 conroe core 2 in the bogomips benchmark."
By that logic, my P4 920@3.8GHz beats them all by having ~8000 bogomips. My current work PC with 3GHz Prescott has 6145.49 bogomips.
Btw, what do you think the "bogo" part means in that unit?
You can't really rely on BogoMips. The BogoMips value gives some indication of the processor speed, but it is not very scientific. Relying on bogomips is the same as relying on SuperPi. None of them really means anything in the real world. What are really needed are new benchmarks that take the most used applications (ex. Photoshop, Excel, Oracle, etc.) and use them to run tests by simulating a user's action (mouse clicks, selecting a load, performing certains actions, etc.). Of course, setting up such an environment is not trivial, but only something like this would give you a real indication of what real performance is like on these processors. :)
I can confirm Sharikou's findings. I purchased an Intel e6400 system specifically for testing. When my brand new Intel e6400 with 1GB ram is put up against my aging AMD 2200+ with 1GB ram, there is little noticeable difference. It is clear that there are serious problems with Intel's entire platform that go beyond the flawed Core 2 architecture.
"I can confirm Sharikou's findings. I purchased an Intel e6400 system specifically for testing. When my brand new Intel e6400 with 1GB ram is put up against my aging AMD 2200+ with 1GB ram, there is little noticeable difference. It is clear that there are serious problems with Intel's entire platform that go beyond the flawed Core 2 architecture."
This clown must be starring at the mobo and seeing all the same components, one cpu, two memory sticks, a hard drive and a bunch of wires....YUP NOT MUCH NOTICEABLE DIFFERENCE HERE... People like that guy need to leave the puters alone and take up NASCAR, you heer!!
It is clear that you intel fanboys dont understand how your windows or linux programs determine the execution speeds.
Bogomips is a Linus Torvalds invention that duplicates a similar function in windows.
Upon os(operating system) start up(when you push the on button to start your computer)the bogomips are calculated.
Bogomips is a calculation based on the program asking the cpu for information.
That calculation renders the highest speed at which the cpu can process nothing(very appropriate for the conroe).
This speed than determines how fast the dynamic timers will be set which directly affects the rate at which the operating system will run.
The higher the bogomips figure the smaller time delays will be set in the program and the faster it will run.
In the example I gave you that means that the AM2-5000 can handle data and process faster than the e6600 conroe these are simply the facts and science that run your computers everyday.
You know you have computers and you could access the internet and get a little smarter?
You intel fanboys should really understand some basic computer operating system functions.
Maybe if you tried a little programing you might understand your computers a little.
Like science mark benchmarks, what are they and why does AMD always beat out Intel on science mark?
"you sir are a imbecile" The data suggests that I am not a imbecile.
A p4 920 is a 2.8ghz. chip I already tested that one and it does not have 8000 bogomips but unlike the conroe it can run basic code without errors.
"The higher the bogomips figure the smaller time delays will be set in the program and the faster it will run."
According to my personal experience the bogomips have almost nothing to do with responsiveness. A 500MHz P3 had several times lower bogomips than dualcore P4 at 8x the clockspeed per core but desktop responsiveness were almost the same. Though I'm not sure how they would have behaved under Windows, that OS can't use disk cache efficiently.
"Maybe if you tried a little programing you might understand your computers a little."
I've tried, about 8 years or a bit less than third of my life and I think I know a few things about CPU's and common performance bottlenecks.
One thing I know is that C2D is the CPU for doing SIMD calculations beating anything else under the sun by at least 2x. I hope K8L is able to be even better than that though it will not be an easy task.
"A p4 920 is a 2.8ghz. chip I already tested that one and it does not have 8000 bogomips but unlike the conroe it can run basic code without errors."
Yes, by default it runs at 2.8GHz but I had mine running @3.8GHz and then it had ~9k bogomips per core.
Forgot to mention that my current C2D can run any code I've trhown at it with nice performance, even when OC'd >70% with just air cooling.
If you know anything that can't run on C2D please tell me what it is so I could verify. You surely seem to know something since otherwise you wouldn't make such claims.
Sharikou you coward! Join Tomshardware forums.. seek out ElMoIsEviL and let's have a debate. I've asked you many times but you continuously ignore me. Are you chicken or something?
Some crap spewing moron said...
"Sharikou you coward! Join Tomshardware forums.. seek out ElMoIsEviL and let's have a debate. I've asked you many times but you continuously ignore me. Are you chicken or something?"
Haha! A debate? On THG forums??? You're surely smoking something pal, that place is a KNOWN Intel fanboy forum/site. I laugh till my sides hurt at some of the crap that is spewed there as "facts", that site and that forum are a farce and nothing more than a front for Intel's marketing machine. The sad fact is MOST people out on the net know it too! Sharikou and most knowledgeable people(read: who are not fanatical fanboys) know better than to waste their time with the mouthbreathers who infest that site and do nothing but "sing praises of Intel". Everyone knows Tom Pabst and THG is a paid Intel pusher cleverly disguised as an excuse for a "online computer hardware review site"
So in short, here's a clue for you:
GO PISS OFF and go rejoin your pals on THG in the Intel circle jerk. Leave the actual debates to those of here who actually know somewhat what they're talking about and don't praise Intel for everything under the sun and the moon.
YOU DON'T LIKE THE BLOG OR IT'S OWNER'S OPINIONS? HERE'S A QUARTER, " © " GO CALL SOMEONE WHO CARES! HIT THE ROAD AND START YOUR OWN BLOG! THIS GOES FOR ALL OF YOU SHARIKOU HATERS!
Hi I'm a fan of AMD/ATI/Linux but I'm willing to admit that the Core 2 Duo is fast/cost effective and that using Windows isn't very painful.
Why are you all making such a big deal out of which one runs faster? There isn't really a whole lot of noticeable difference.
Sometimes one side is winning, sometimes the other side is.
What's so hard to accept about that? Why are you all getting so stressed out over it? o_O
Why do you call somebody a fanboy of something just because they will admit that it works good? Come on guys it's just a little piece of silicon..
I bought a core 2 duo E6750, 2.66 Ghz FSB1333, using 2Gb of 800Mhz DDR2 RAM. after upgrading from my Xeon 1.8Ghz processor that I had before, I am EXTREMELY disappointed in the POOR response. I can see every window or click response. This is one crap that Intel has come out with I don't know whether to blame Intel or Microsoft's Windows XP SP2 but they together is the greatest disappointment in an upgrade I've ever done in my 20 years of computer experience. I want back my money !!!
OK i I have to post because i have tried the C2D experience and I am not an typical C2d fan boi. it has not been any where close to worth is money and I'm sick of ppl saying it worth the xtra dosh. a years worth of of disapointment. I'm now an avid AMD supporter.
Post a Comment
<< Home