Whats the difference between the X2s aside from clocks?

S

Svengali

Limp Gawd
Joined
Mar 13, 2006
Messages
207
Lets say you have a FX-60 (939) @ 2.6, and OC it somehow to 3GHZ? Is it the same as a 6000+?

I mean, isn't the architechture really the same, it appears to me its just clock differences on this line. Cache as far as I know doesnt make much if any difference in gaming (which is why im asking)
 
If running under the same conditions (i.e. both are using the same type of RAM, same cache, same frequency) they will perform identically. 65nm brisbane core processors have slightly higher cache latency than their 90nm brethren, but it won't make a perceptible difference in real-world use. As you mentioned, cache doesn't have a significant effect on performance on K8s.

One main difference between the FX line and the standard X2 is the advantage of having a fully unlocked multiplier, which makes overclocking very easy. The X2s can only drop their multiplier, and cannot increase it.
 
OK I've just been wondering since im on a X2 4800 (2.4ghz) and I'm surprised the best AMD has to offer is simply what amounts to a 600mhz overclock. Do you know what the X2 4800+ is comparable to a Core 2? In terms of real world gaming performance.
 
Gaming performance is very similar and will generally be nothing more than minor frame rate differences.

[H] did a cpu gaming article and a 8800GTX SLI one comparing cpus. The FX-62 (2.8GHz) gave roughly the same gaming performance as the x6800 (2.92GHz) in gaming. In the SLI one, the Intel setup had a bit better shadows in Oblivion and a bit higher AA in Flight Sim X.

Now in a pure cpu power test like encoding or something an e6600 (2.4GHz) would beat that FX-60. But with gaming it's mainly just some fps here and there.

If your gaming performance is lagging, upgrade your graphics card as that'll give much more of a boost than a faster cpu.
 
X2 (Manchester core) 2x512k L2 Cache, 90nm, socket 939, DDR1 memory controller locked multipliers
X2 (Toledo core) 2x1024k L2 Cache, 90nm, socket 939, DDR1 memory controller, locked multipliers
FX-60 (Toledo core) 2x1024k L2 Cache, 90nm, socket 939, DDR1 memory Controller, unlocked multiplier

X2 (Windsor core) 2x512k or 2x1024k L2 cache, 90nm socket AM2, DDR2 memory controller, locked multiplier
X2 (Brisbane core) 2x512k L2 cache, 65nm, socket AM2, DDR2 memory controller unlocked multiplier
FX-62 (Windsor core) 2x1024k L2 cache, 90nm, socket AM2, DDR2 memory controller, unlocked multiplier

FX-70, FX-72, & FX-74 (Windsor core) 2x1024k L2 cache per socket, 90nm, socket 1207-FX DDR2 memory controller, unlocked multiplier.

Other than those differences they're just dual San Diegos with minor tweaks (Toledo) and different memory controllers (Windsor/Brisbane).
 
So then it wouldnt be worth it to try an OC my 4800 any would it? I dunno whats with it it idles at like 47-50C with a Arctic Freezer 64 on it. Seems pretty hot (939 4800) I was thinking if I got a 8800 it would be bottlenecked a bit and a lil OC might relieve that but maybe not, eh.

Does anyone know if theres a good direct comparison with the 4800 to a Core2 Like e6300 etc..
 
Svengali said:
So then it wouldnt be worth it to try an OC my 4800 any would it? I dunno whats with it it idles at like 47-50C with a Arctic Freezer 64 on it. Seems pretty hot (939 4800) I was thinking if I got a 8800 it would be bottlenecked a bit and a lil OC might relieve that but maybe not, eh.

Does anyone know if theres a good direct comparison with the 4800 to a Core2 Like e6300 etc..
Click to expand...

Worth is in the eye on the beholder. Overclocking is entirely up to you. You might see some gains, you might not notice them. As far as the temps, how much airflow do you have and what is the temp in the room like. Those two things can play havoc with CPU temps.
 
Dunno about room temp, it goes down a good bit in the winter but it really isnt optimal by my best guess. Airflow is really good its a really well wired Lian Li case (Velocity Micro) I dust the case out and clean the fans regularly as well. Theres no crap in the way of anything, I am running 7900GTs in SLI maybe thats increasing the case temp.

As for worth, I'm just worried that I'll have to run lower settings than someone on a comparable configuration to mine, with a better CPU I.E. A 8800GTX on a E6600 vs on my 4800. Do any of you think the difference in performance will be enough to limit the settings in games out or coming out? I'm typically fine with 2-4X AA and any AF I can get, but I do like to run the max in game settings, and If I spend that much on a new graphics setup I'd like to be able to take advantage of that. It's just been fairly hard to find a definitive answer on whether or not my CPU will hold me back (I'll be playing at 1600x1200)

I mean if the difference gap isnt that big anyways why'd everyone jump ship to Intel for gaming? I understand why for other purposes, but for gaming a lot of the time I hear it won't make much of a difference but its hard for me to believe I guess. I mean I've seen those benchmarks referenced above and others that tend to prove it doesnt have much impact, but then I'lll see benchmarks whose goal isn't to prove whether or not CPU has a significant effect on realworld performance, and it turns out in those benchmarks it seems to be significantly different. Many times as much as 20-25fps. While a comparable PC is getting 90, and the top end cpu one is getting 120 or something like that, obviously thats moot as 90 is more than enough

. But when it comes to stuff like BioShock or Crysis I worry it won't be moot differences. It'll be the difference between playability and unplayable at those settings.

Dunno if that makes any sense.
 
The numbers are still "moot" because it's not a hard value, but a ratio.

If going from an AMD X2 '4800+' to an Intel Core 2 Quad bumps you from 120fps in a game to 150fps...yeah, that's an improvement of 30fps. But that doesn't mean you would be otherwise looking at 15fps for the AMD or 45fps for the C2Q in a high-intensity game. 120 -> 150 is only a 20% improvement. If the game was running at 15fps on the AMD, it'd only be up to 18fps on the C2Q.

I think we can agree that "moot" is "moot".

Unless you are gaming at 800x600 with no AA and no AF settings, I've never seen a performance difference between relatively equiv price-point-at-release CPUs top 25%, unless you are separating them by many generations. (The only real way I see this changing is in architectural differences. Clock speeds aside, a single-core CPU just isn't going to hack it on a game that requires two cores, and benefits from 4. And a dual-core CPU is going to have a rough time of a game that requires 4 cores.)

9 times out of 10, as the [H] article indicated, you are ALWAYS better off putting that money into a better video card for PC gaming.
 
Svengali said:
As for worth, I'm just worried that I'll have to run lower settings than someone on a comparable configuration to mine, with a better CPU I.E. A 8800GTX on a E6600 vs on my 4800. Do any of you think the difference in performance will be enough to limit the settings in games out or coming out? I'm typically fine with 2-4X AA and any AF I can get, but I do like to run the max in game settings, and If I spend that much on a new graphics setup I'd like to be able to take advantage of that. It's just been fairly hard to find a definitive answer on whether or not my CPU will hold me back (I'll be playing at 1600x1200)
Click to expand...

http://www.hardocp.com/article.html?art=MTI2MiwxLCxobmV3cw==

http://www.hardocp.com/article.html?art=MTEwOCwxLCxobmV3cw==

Those are the articles they did with cpu's and gaming performance. One thing to remember about newer games is that they will be using more than one core, which should make clock speed per core less important as the work will be spread out.


As for the switching to Core2 chips, it's the enthusiast thing to do. Whatever has the most power when you upgrade is what to go for. And many people see it as an excuse to upgrade and do so. After doing the upgrade myself, I have to say I wish I stuck with my A64 system as I didn't get any kind of real in game performance upgrade out of switching. Nothing big and noticeable at least.

EDIT:: Wanted to add a quote from the 8800GTX SLI review that seemed like it would help you.

There is undoubtedly a world where the Intel Core 2 Duo reigns supreme, but in the world of gaming that power is not near as apparent. We again see games remain very GPU dependent, but the GeForce 8800 series of GPU is allowing us to begin to see the Core 2 Duo flex its muscle. One thing is for sure. If you are building a gaming system, your big money is better spent on your video cards than your processor as you are never likely to see any true differences in your gaming experiences without a hugely powerful GPU.
Click to expand...
 
You must log in or register to reply here.
Back
Top