• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

[email protected] VS [email protected] With 4GB of Memory

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

f012t12

Member
Joined
Dec 7, 2003
Location
New Jersey
Which core would best use the memory if an application were demanding it?

Say the two systems based on these two cores had 4x1GB@?@2-3-2 timings?
 
Athlon 64.
P4 is going to throttle down unless you have killer cooling.
Just look at any review that takes throttling in effect from the 3.6 or 3.8.
With stock voltage at stock speed, it doesnt run at ful speed very often.
 
Well, they are both dual channel (and at the same speed here), but the AMD processor gets the advantage because it's on-die controller reduces memory latency.
 
dicecca112 said:
I think another relevant question is what are you doing with these chips that you need 4gb of memory?

Well, what's the point of ever having a 64-bit compatible CPU if you can only allocate a maximum of 4GB of memory?

So, if 4GB is the limit, what product can make use of it most?
 
f012t12 said:
Well, what's the point of ever having a 64-bit compatible CPU if you can only allocate a maximum of 4GB of memory?

So, if 4GB is the limit, what product can make use of it most?

4Gb is not the limit for a K8-based processor. If you can't find a desktop board that supports more, try a server board I know dual boards allow more. Now, Xeons can work around te 4G limit (I believe), but the K8 can directly address (no workarounds, uses 48-bit memory addresses) 1TB of Physical memory and 256TB of virtual memory, when running in 64-bit long mode (sandpile.org).

edit: There are two main reasons to move to x86-64 other than the memory limit. The first is that you can handle 64-bit integers much faster, but most people don't use those anyway. The second is hat you have 8 more general purpose registers and 8 more SSE2 registers. Since x86 is a fairly register-limited arch, this can provide a performance improvement by justt re-compiling code for the new arch to make use of the new registers. There is another non-64 bit main advantage of the K8, which is the on-die memeory controller. This reduces the latency of main memory.
 
Last edited:
The P4 and A64 will both run with lower timings if you have more then 2 sticks of ram on the motherboard. Also the P4 will disable PAT. There was a test about it some while ago, and while everything that did'nt have registered ram performed worse (when the extra ram is not in use) the P4 degraded the most and A64 the least.
 
dropadrop said:
The P4 and A64 will both run with lower timings if you have more then 2 sticks of ram on the motherboard. Also the P4 will disable PAT. There was a test about it some while ago, and while everything that did'nt have registered ram performed worse (when the extra ram is not in use) the P4 degraded the most and A64 the least.
I did not mean that the A64 could run the memory at lower latency, I simply meant that, all memory timings being the same, the A64 will have lower latency for memory acces because of the on-die memory controller.
 
the fx55 is gonna be miles ahead. even stock its gonna crush a 4GHz p4 just silly. alot of mobos may not handle 4 gigs of ram, also having this much ram will really hurt overclocking and performance
 
Gnufsh said:
I did not mean that the A64 could run the memory at lower latency, I simply meant that, all memory timings being the same, the A64 will have lower latency for memory acces because of the on-die memory controller.

I was not aiming the post at you, actually you where completly correct. The point I was trying to make was, that unless he is really using all that ram he is better off running the computer with 2 sticks of ram. :)
 
Back