Hi,
I’ve been Googling quite a while now and I figured I’d come here and just ask to see if there is a “resident geek” that could answer my question about how AMD processors work. I’ve always been an “intel person” you can easily find the FSB speed with these and match an appropriate MHz stick of ram to know there are no bottle-necks there.
But with AMD processors there are certain multipliers happening? which makes it hard to do the same procedure. The processor in question is a AMD Athlon 64 X2 tk55… it’s stating the FSB is 1.8Ghz… well this computer is not very new so I know there’s some math going on to get that number. Would anyone know how to figure out the best suited speed of memory that matches the FSB on this?
Yes, I’ve thought about getting faster ram and letting the system clock the speed back for me… but it’s kind of an opportunity for me to learn more about AMD… so any ideas?
they don’t have a traditional FSB, as the memory controller is integrated into the CPU. The CPU and motherboard documentation will tell you the type of memory to use.
By the way, the latest Intel CPUs do this also. Front-side buses are so 2006.
This is something that mostly overclockers tend to worry about. I would check out some of the more reputable computer forums like XtremeSystems. In addition to the speed there are also other considerations such as strappings, ganged/unganged, etc. And I don’t know enough about such things to advise you. You will find more info than you can handle at sites such as the one I mentioned.
Hm pretty sure you can use a variety of memory speeds with processors these days…last i checked, the memory clock is independent from cpu clock. even with “appropriate” mhz sticks come with different timings based on what you are willing to pay. In any case memory is a minor issue at this point, the bottlenecks are actually at storage, which is why ssd’s are so popular. The only issue is type of socket and mb support, laptops are the pickiest.
From what I recall, you do actually get better performance by matching the bus speed to the memory speed but I can’t recall all of the reasons. I think it has to do with limiting the number of hard faults, but IDK.
You’re right that disk tends to be a bigger problem, but I wouldn’t say that the memory bottleneck is “minor”. The memory bottleneck is getting worse as processors improve and cores are added, which is why there is so much L1, L2 and L3 cache being added on-chip.
Well relative to performance increase in cpu yes, but as a matter of which subsystem holds the system back the most…its not lol. Dan of dansdata had an article on the memory thing a long while back.