Intel q6600 vs AMD Phenom. I am betting with a friend

Status
Not open for further replies.
I had the same kind of bet, but we ran a battery of benchmarks and allowed air cooled OCing.

Bottom line is that core2's OC way too well most of the time to be beaten if it's allowed. Stock clock for clock AMD is a good value still.
 
kassler, you're a comedy genius: please, do continue - my afternoon has been significantly enriched by your 'knowledge' :D Do you offer 'hole digging' courses??
 
jellyhead said:
kassler, you're a comedy genius: please, do continue - my afternoon has been significantly enriched by your 'knowledge' :D Do you offer 'hole digging' courses??
Click to expand...
Stupid people find knowledge funny or scary. They do react strangely. What they can’t do is to talk about it.
 
kassler said:
Can you find a review on a more trustworthy site? I hope you know that anandtech is VERY PRO Intel (could be some economics involved…)
Click to expand...

Why don't you run cache2cache for yourself? You'll see that you get the same results as Anandtech. Or just use google for more results.
 
I got the following result on a stock clocked Q6600. I do have a quad core xeon server here I could run it on later, but that's only an E5335.

I found it a bit strange that the random results were so low. No matter how many times I run it, the result is the same. Interesting.

q6600cache2cache.jpg
 
MrMike said:
I got the following result on a stock clocked Q6600. I do have a quad core xeon server here I could run it on later, but that's only an E5335.

I found it a bit strange that the random results were so low. No matter how many times I run it, the result is the same. Interesting.
Click to expand...

The cache (hardware design) on intel is sensitive to fragmentation. It is fast when things are linear.
 
kassler said:
The cache (hardware design) on intel is sensitive to fragmentation. It is fast when things are linear.
Click to expand...

I think you misunderstood. By low I meant fast. The random results I got are extremely fast. They're 1/3rd of what Anandtech was showing.
 
MrMike said:
I think you misunderstood. By low I meant fast. The random results I got are extremely fast. They're 1/3rd of what Anandtech was showing.
Click to expand...

Anandtech says: "The results that we publish are approximately twice the propagation time"

So they doubled the readings in their table, because they wanted show how long it takes for data to travel to another core and back.
Even so, your CPU is still faster... But you have 1066 FSB and 2.4 GHz clock, theirs has 1333 FSB and 2.33 GHz clock. I wonder if the Xeon chipset/motherboard has any effect on it (being an FB-DIMM setup with 2 sockets).
The L1 speed seems about the same on the Xeon.
At any rate, it doesn't look like Anandtech is trying to fudge the cache2cache results in Intel's favour.
 
kassler said:
The cache (hardware design) on intel is sensitive to fragmentation. It is fast when things are linear.
Click to expand...

Fragmentation?
Hah... If you are trying to refer to the associativity, it is well higher on Intel CPUs.
 
kassler said:
Can you find a review on a more trustworthy site? I hope you know that anandtech is VERY PRO Intel (could be some economics involved…)
Click to expand...

I seem to recall them pushing whatever was the better performer at the time. When AMD Athlon64 (X2) and Opteron chips were the best, they were very PRO AMD. When Core architecture came out, they became more PRO Intel. I don't think they're biased, but thats just me and what I've seen over the years. Their build guides always have a build for each type of system, Intel or AMD, too. I think they're less biased than THG, and I trust them more than THG.
 
HAHA this has been entertaining :) about 4 to 5 people that know what they are talking about and one person that doesn't have a clue :) :) its great entertainment no doubt :p
 
this is a interesting thread. I though the debate for what was better now was over. I got a x2 3800+ when all my buddies got a e6600. I tried to tough it out but it just didn't work out. AMD had a good chip not a great chip especial when compared to what Intel has out now. You can look at the good parts of what amd can do better but when the some of the parts are put together and one comes out on top that where I'll put my money.
 
On a stock 3.0 GHz Core 2 Duo E6850, I get:

20hlfrs.png


Very fast, as expected. Elios should take care of the 45nm Core 2 Duo results. Now we just need some 45nm Core 2 Quad and AMD results. :)
 
hmm thats interesting the 45nm at lest mine takes a few more cycles but over all speed is faster and higher bandwidth

wonder if that is do to only 3MB of L2?
 
I'm going to make my own company and start making cpu's which will blow even intel out of the water! goign to call the 1st line THisshitisfastthanintelaswellasAMDandisbetterpriced. I'll so be the talk of the town
 
lies! you're all intel conspirators posting fake numbers.

reminds me of the iraq minister of information guy during the initial invasion
 
Your friend does seem really confident. He also seems really dumb. Are you telling me that there are THOUSANDS of sites, all of which are endorsing Intel? Tell him to get real, I say.

But, this one is the major deciding factor. Do you prefer him as a friend, or do you prefer to win :p. I wanna see somebody feel stupid, so please, do win it :D.
 
simple ... run supi .... newest amd quad gets 25 secs on 1m ... my laptop dual gets 22.225
 
I just hope you two are still friends after this betting no matter who wins it. Intel will win, I believe.
 
kassler said:
I think you will find this text interesting
Studies of Threading Successes in Popular PC Games and Engines
Click to expand...

Not really, been there, done that ages ago (and the test application I linked earlier is testament to that, much better multithreaded scaling than most games, it pretty much implements everything mentioned in that article). As I say, I have over 20 years of experience.
I don't see how it is relevant to the post you quoted either.
 
Are you going to post cache2cache results from your Phenom Kassler? We've given you "non-Anandtech" Intel numbers like you wanted.
 
To update with you guys: My friend cheated a bit. He bought the CrossHair 2 Formula motherboard which is much more than my Asus P5K-E. He bought the Phenom 9850 Black Edition, $200. And he bought the Corsair PC8500 8GB RAM which costed twice as much as my cheap 8GB Patriot PC6400 RAM.

He said as long as the CPU cost the same as the Q6600.

Yesterday he tried to install Vista 64bit Ultimate and it crashed on him during the install. Later I told him it's because he is supposed to have only 1 memory module 2GB during the install, and intall the rest afterwards.

Oh, he is going to use stock cooler. I do hope so. But I think he will wise up and go for a big cooler soon. I told him yesterday: "Unless you overclock it, you have no chance against me."
 
adobian said:
To update with you guys: My friend cheated a bit. He bought the CrossHair 2 Formula motherboard which is much more than my Asus P5K-E. He bought the Phenom 9850 Black Edition, $200. And he bought the Corsair PC8500 8GB RAM which costed twice as much as my cheap 8GB Patriot PC6400 RAM.

He said as long as the CPU cost the same as the Q6600.

Yesterday he tried to install Vista 64bit Ultimate and it crashed on him during the install. Later I told him it's because he is supposed to have only 1 memory module 2GB during the install, and intall the rest afterwards.

Oh, he is going to use stock cooler. I do hope so. But I think he will wise up and go for a big cooler soon. I told him yesterday: "Unless you overclock it, you have no chance against me."
Click to expand...

At stock clock , put the RAM on 1066 CL5.
 
He's hosed either way. Yours is typically faster stock and they OC better.

In synthetic benchmarks it's hard to tell who will win all the time, but I'm pretty confident the C2Q will do well.
 
adobian said:
To update with you guys: My friend cheated a bit. He bought the CrossHair 2 Formula motherboard which is much more than my Asus P5K-E. He bought the Phenom 9850 Black Edition, $200. And he bought the Corsair PC8500 8GB RAM which costed twice as much as my cheap 8GB Patriot PC6400 RAM.

He said as long as the CPU cost the same as the Q6600.

Yesterday he tried to install Vista 64bit Ultimate and it crashed on him during the install. Later I told him it's because he is supposed to have only 1 memory module 2GB during the install, and intall the rest afterwards.

Oh, he is going to use stock cooler. I do hope so. But I think he will wise up and go for a big cooler soon. I told him yesterday: "Unless you overclock it, you have no chance against me."
Click to expand...

i had all 8gb of my ram in for the install and no problems. maybe ur friend doesn't know what he is doing.
 
adobian said:
To update with you guys: My friend cheated a bit. He bought the CrossHair 2 Formula motherboard which is much more than my Asus P5K-E. He bought the Phenom 9850 Black Edition, $200. And he bought the Corsair PC8500 8GB RAM which costed twice as much as my cheap 8GB Patriot PC6400 RAM.

He said as long as the CPU cost the same as the Q6600.

Yesterday he tried to install Vista 64bit Ultimate and it crashed on him during the install. Later I told him it's because he is supposed to have only 1 memory module 2GB during the install, and intall the rest afterwards.

Oh, he is going to use stock cooler. I do hope so. But I think he will wise up and go for a big cooler soon. I told him yesterday: "Unless you overclock it, you have no chance against me."
Click to expand...

He'll have no chance against you overclocked. He won't get past 3.2GHz and you should be able to do that just as easily. Even if you only hit 3.0GHz you'd be about the same speed in most applications. He'll have to have about 400MHz on you or more to pull off a general victory. That's not likely to happen.
 
Dan_D said:
He'll have no chance against you overclocked. He won't get past 3.2GHz and you should be able to do that just as easily. Even if you only hit 3.0GHz you'd be about the same speed in most applications. He'll have to have about 400MHz on you or more to pull off a general victory. That's not likely to happen.
Click to expand...


Thanks, this sure is nice to know.

Anyway, does anyone recommend turning off the page memory? I have 8GB of RAM. Will this boost the performance? Probably not when you have to RIP two Blue ray discs at the same time ? I saw my friend with a new external Blue Ray Drive. He is up to something.
 
No, the system will perform fine with a page file, windows memory manager isn't perfect but it isn't retarded either.

Ripping two blue ray discs at the same time is a pretty silly benchmark.

Also if you're using external drives, the scores will be the same because it will be slowed down by the link you're using. Hell, even internal drives would be limited by the speed of the interface and disc read speed more than likely.
 
it has nothing to do with cores or clock speeds, the core 2 architecture is just faster/better design then the AMD. I used to be all AMD till about 8 months ago when I finally tried an intel. They just have better chips then AMD.
 
You should bench it with Vdubmod encoding something Vdubmod supports both amd and intel AND multiple cores :)
 
Volcanon said:
Ripping two blue ray discs at the same time is a pretty silly benchmark.
Click to expand...

I don't know, that sounds to me like something that would demonstrate good real world performance rather then a static benchmark that outputs a number that you can put up on a chart on your wall. If that's what he will be doing a lot of then why not have it in the "benchmark suite."
 
Status
Not open for further replies.
Back
Top