- Joined
- Oct 17, 2000
- Messages
- 1,875
How does Conroes 64bit match up to AMD's 64bit?
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
-
Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.
C2D 64bit as good as AMD 64bit?
- Thread starter dave343
- Start date
Black Morty Rackham
Supreme [H]ardness
- Joined
- Jan 8, 2004
- Messages
- 4,728
Eh, it's just 64-bittiness. It's pretty much an either-or kind of thing. Also, unless you need more than 3 gigs of RAM, you shouldn't really care.
I believe that the gap between conroe and AMD is smaller under 64-bit, but conroe still generally performs better for the money (without factoring amd's pricecuts into it). There haven't been a whole load of comparison articles, but if you search around you should be able to find a couple.
C2D is still faster and costs less in most cases.
bubblescivic
Limp Gawd
- Joined
- Aug 17, 2006
- Messages
- 163
64 bit instruction is 64 bit instruction regardless of brand. however, C2D kicks AMD's ass left and right.
Can we find some actual benchmarks on this topic?
These are old benchmarks of P4 vs Athlon 64, but that doesn't change what they show. Intel's processors performing worse in 64-bit than 32-bit, while AMD's perform better.
Does this same trend continue with C2D? If so, then the answer to the question of the thread is no, AMD's implementation of 64-bit is superior. Lets try and stay on topic and not spam this thread with 'OMG C2D ALWAYS FASTER!!!!!!11111111111111' please. I am also very interested in seeing which processor performs better in 64-bit vs their performance in 32-bit. I can honestly say that 64-bit WinXP feels faster than 32-bit WinXP on the same AMD64 machine, but I have no C2Ds to test.
Edit:
This thread here at AMDZone proceeds to suggest that with large amounts of memory, the Opterons are noticeably superior to Xeons. This would more than suggest that Intels EM64T is inferior, but why? Most likely the fact that the Opteron has the on chip memory controller.
These are old benchmarks of P4 vs Athlon 64, but that doesn't change what they show. Intel's processors performing worse in 64-bit than 32-bit, while AMD's perform better.
Does this same trend continue with C2D? If so, then the answer to the question of the thread is no, AMD's implementation of 64-bit is superior. Lets try and stay on topic and not spam this thread with 'OMG C2D ALWAYS FASTER!!!!!!11111111111111' please. I am also very interested in seeing which processor performs better in 64-bit vs their performance in 32-bit. I can honestly say that 64-bit WinXP feels faster than 32-bit WinXP on the same AMD64 machine, but I have no C2Ds to test.
Edit:
This thread here at AMDZone proceeds to suggest that with large amounts of memory, the Opterons are noticeably superior to Xeons. This would more than suggest that Intels EM64T is inferior, but why? Most likely the fact that the Opteron has the on chip memory controller.
Knowledge talks , BS walks...
http://www.xbitlabs.com/articles/cpu/display/core2duo-64bit.html
There is no speed loss due to EMT64 vs. AMD64.It is simply BS made by !!!!!!s of you know who , desperate to find something to grasp for when their ideal world crumbled.
At the end the conclusion is flawed because the molecular test is disruptive and everyone who studied a little bit of maths knows that you have to remove the anomalies in order to get the median.In a similar sens , if calculating the speed advantage of Conroe vs. K8 you need to remove SuperPI because it will alter the results.
The conclusion : it is a matter of implementation ; no side has the upper hand.In some tests Conroe gains more % wise , in some K8.Overall , P4 with EMT64 gains most , Netburst rules!
* A special mention for Open Source SW.
As you know the OS community was the 1st to adopt 64bit and because Intel was 1 year late to the party ( spring 2003 vs. summer 2004 ) all the tools( especially compilers , GCC that is ) and optimizations for K8 only.As a result open source SW gains most from 64bit on K8 cpus.The situation is slowly improving as more and more developers optimize for EM64T on Netburst/Core CPUs.
http://www.xbitlabs.com/articles/cpu/display/core2duo-64bit.html
There is no speed loss due to EMT64 vs. AMD64.It is simply BS made by !!!!!!s of you know who , desperate to find something to grasp for when their ideal world crumbled.
At the end the conclusion is flawed because the molecular test is disruptive and everyone who studied a little bit of maths knows that you have to remove the anomalies in order to get the median.In a similar sens , if calculating the speed advantage of Conroe vs. K8 you need to remove SuperPI because it will alter the results.
The conclusion : it is a matter of implementation ; no side has the upper hand.In some tests Conroe gains more % wise , in some K8.Overall , P4 with EMT64 gains most , Netburst rules!
* A special mention for Open Source SW.
As you know the OS community was the 1st to adopt 64bit and because Intel was 1 year late to the party ( spring 2003 vs. summer 2004 ) all the tools( especially compilers , GCC that is ) and optimizations for K8 only.As a result open source SW gains most from 64bit on K8 cpus.The situation is slowly improving as more and more developers optimize for EM64T on Netburst/Core CPUs.
This is exactly what I was looking for, thanks. The gains in most things really are about even. So basically, unless you're going to have absurd amounts of memory, it doesn't matter. I would imagine very, very few members here have 16GB+ RAM, and in systems of that scale I am personally more concerned about the support and warranty that is included with it rather than negligible seconds.
Indeed, which is one of the reasons I was looking for something like the benchmark you posted.
RAM quantity has nothing to do with how much you gain from 64bit.It all boils down to architecturals tradeoffs inside the CPU.
There are apps where AMD does see a bigger gain. The problem is it ends up getting beat by 28% in 64bit instead of 32% mode in 32bit mode
Then there are other apps where it looses ground. As was pointed out already hardly any of the software is opimized for C2D and it is still kicking ass.
That's what I was saying, just reinforcing my first post on that point. That an AMD chip will still perform better with exceptionally large amounts of RAM than an Intel chip because of the integrated memory controller. Those large amounts of RAM aren't an issue at all in 32-bit computing because they do not exist.
The integrated controller only reduces latency.
The amount of memory doesn't matter as far as the memory controller is concerned.
It just opens and closes memory pages, and initiates transfers. It doesn't even know how many pages there are. It just does what it's told.
As long as you have 'enough' memory for what you're doing, it doesn't matter whether your system has eg 1 gb, 2 gb or 64 gb. The access time and memory transfer speed remain exactly the same (assuming identical memory and no shortcomings of the controller or motherboard).
A large amount of memory may only mean that the OS has a tougher job managing it (but generally they are O(1) algos, so even there you don't notice), but since both Intel and AMD use the same instructionset and the same OSes, there is no difference between architectures.
The amount of memory doesn't matter as far as the memory controller is concerned.
It just opens and closes memory pages, and initiates transfers. It doesn't even know how many pages there are. It just does what it's told.
As long as you have 'enough' memory for what you're doing, it doesn't matter whether your system has eg 1 gb, 2 gb or 64 gb. The access time and memory transfer speed remain exactly the same (assuming identical memory and no shortcomings of the controller or motherboard).
A large amount of memory may only mean that the OS has a tougher job managing it (but generally they are O(1) algos, so even there you don't notice), but since both Intel and AMD use the same instructionset and the same OSes, there is no difference between architectures.