Basically I put this configuration together under 300 £.
I previously only had a
HD 6950 2gb
rest I don't remember.
I thought it was a good and relatively cheap upgrade. Now my friend tells me
that the processor was not really an upgrade as the maximum clockspeed of the i5 2500k is around the same as the q6600. It is true that i can still overclock the i5 but is the q6600 at 3.6 really as fast as the i5 2500k at 3.6?
If not what factors does the speed really depend on?
Clock speed is not the determining factor in CPU performance, there are many differences in architecture. Although Sandy Bridge (like that i5) is a little old, it will catch up to a new i5 4670K with an overclock. In any case, it will perform worlds better than that old Core 2 Quad, especially when paired with an R9 280x.
Your friend is totally wrong, if every CPU clocked at 3.6GHZ functioned the same, why would exist the need to sell Xeons, or an I7? If it worked along your friend's logic, even a Pentium 4 at 3.6 GHZ could match the processing capacity of the i5 2500k. Don't believe him.
Each new CPU architecture improves IPC (Instructions per clock) that is, the processor can process more information for the same quantity of time. That's what makes CPU's have differences between each other. This is why some AMD Cpu's at 5 GHZ are still slower than 3GHZ Intel CPU's for example, because Intel's Cpu's have a higher IPC.
You can't use the clock speeds as a comparison between 2 cpu's of different architecture. The I5 is capable of processing more instructions per clock, so even at lower clock speeds it would still be faster than a core 2 quad.
The way to think of all this Instruction Per Clock (IPC) stuff is by an analogy to spinning gears turning a wheel. If you spin a very small gear around once, how much will it spin the wheel? Spinning a gear of one size might spin the wheel only 1/4 turn, but spinning a gear of another size might spin the same wheel two full turns. If you spin both gears at the same speed, the wheel still spins much faster on one gear compared to the other. In that example, one gear spins the wheel 8-times faster than the other, even though both gears spin at the same speed. So, even if you spun that gear at half speed, it would still be 4-times faster than the other gear.
Now think of IPC as the gear size; think of "Mhz" as how fast the gear turns; and think of "Floating Point Operations Per Second (FLOPs)" as how fast the wheel spins. Assume for sake of argument that a 4670k is 8 times as fast as the old Pentium 4 was. You'd have the same setup as above. The 4670k @ 2.0 Ghz would be four-times faster than the Pentium 4 @ 4.0 Ghz. That is, in one second you would get four times as many FLOPs from the 4670k as you would the Pentium 4, even though he Pentium 4 has double the clock speed.
There are obviously other variables I left out to keep things simple, and the analogy doesn't hold up perfectly when you factor in multi-core operations and other things. However, you get the basic picture from the analogy. I hope that helps.
That explanation gets at why the CPUs you are comparing are so different despite similar clock speeds. The 2500k is actually multiples faster than the Q6600 according to benchmarks measuring performance. Take a look yourself at this link: http://cpuboss.com/cpus/Intel-Core2-Quad-Q6600-vs-Intel...