Discuss CPU names.. WTF? in the CPUs forum on Dev Shed. CPU names.. WTF? CPUs forum discussing information including overclocking techniques and results. Discuss AMD and Intel based architectures. Disclaimer: Overclocking will void your warranty.
Time spent in forums: 1 Day 19 h 32 m 42 sec
Reputation Power: 26
CPU names.. WTF?
Way back in the days of 2-3-486, it was bloody easy to look at a cpu and say "that one is better than that one." It was freaking clear as day and the numbering was easy to read. This continued up to Pentium 2 generation. After that, the cpu makers got the 'brillant' idea to stop numbeing them by core speed and generation number, and give each gen a name, which to anyone who keeps up with the stuff every month, proly knows the order in which they came out, and what core is better. But for someone like me, who only wants to know that kind of info when i build a new machine, looking at 4-5 new cores all at once for the first time is daunting.
On AMD side we got..
Athlon 64 FX
Athlon 64 X2
Turion 64 X2
Now with the 'Athlons, its kinda readable as to what ones are were in the generation.
Now for intel.
Core 2 Duo
Core 2 Extreme
Core 2 Quad
These are fairly mixed up but this is just the start, then we have to worry about the cores themselves.
I mean.. Wtf? Town names? how the hell am i suppose to make sense of that and tell what one is faster than the other? What happen to the nice easy to read number?
It was easy 'back in the day when you say" "Pentium 2 266 mmx"
You knew exactly what the core was and what speed the dam thing was. Now it seems they are just coming out with new lines like cars. Different models that do different things, but arn't necessarily faster than another. Like "heres a Cavalear, oh also, here is a Sunfire" they look slightly different, and have a different name, but their the same car.
Ive tried looking at Toms hardware (at the cpu chart) but not really sure what the numbers are suppose to represent. But even with that aside, the cores seem to hop around as far as performance. Not the "this core, then this core, then this core" as i kinda expected to see, at least with that i could get a better idea of where they fall in the generation.
Is this a stupid question? (question being, what is the core generation order now??) or am i not alone in thinking this naming thing is stupid and confusing?
Ive poked around on toms hardware and tried googling it, but all i get are places trying to sell me cpus and nothing concrete about where they fall.
Time spent in forums: 1 Month 3 Weeks 3 Days 5 h 12 m 24 sec
Reputation Power: 904
I don't think its that hard at all to understand
On the Intel side you have the fast "Core" based CPUs, and you have the slow Netburst based ones under "Pentium" name. The Mhz can't really be compared between them, put in all but the most extreme cases any Core 2 is faster than any Pentium. Then Intel has the "Celeron" series, they use that name for both types of processors to mean "The garbage that nobody wants", its just crippled and slow. Xeon just means its a "server processor", which adds a better warranty and jumps up the price, its really the same as the desktop though it works in the server sockets as well as the desktop sockets. They also use the 'M' in the name to mean "mobil" (as in its for a laptop). The 'Extreme' CPUs just have an unlocked multiplier (which is only helpful for overclocking)
On the AMD side you have the new K10 core and the old K8 core, pretty much its the same as intel, K10 is much faster then the K8, however with AMD the K10 is only used for their quad core (Phenom). Right now everything else is the K8 core. All the Athlons are pretty much the same right now, they are all K8 CPUs, most of which are dual core. The specs on each one differs a bit too, but its all the same CPU. They use "Sempron" for all their garbage CPUs (just like intels Celeron). Their mobile CPUs are called a "Turion". They also use Opteron for the server processors, again, same deal as the Xeon on the Intel side
For each company the "core" names are just picked by the development team and are really meant for internal use, there is a core for every combination of specs (other then frequency). So two CPUs with the same core have the same size cache(s), it makes its easy if you searching for a CPU with specific features, but otherwise you can ignore the core name.
That pretty much is is, i suppose a lot of the confusion just comes from their cost cutting measures, back with the 486 there CPU was small and they could afford to throw out the bad ones (the transistor count was much lower and had less failure points), now the CPUs are so complex they have to test every one, find what parts are bad, then disable those parts and sell it, thats why you get such the huge variance in available speeds and cache, if it fails a speed test they sell it at a slower clock speed, if it has a bad cache they disable that part of the cache and sell it as having less cache. Thats why the cores don't really matter either, they are all produced as the high end core, and knocked down to the low end ones depending on what they failed, they don't know what core CPU they made until very late it its construction.