How does a 4ghz core 2 quad stack up these days?

[lavaheadache]

hypothetically, what is a 4 ghz Q6600 comparable to these days?

 

[dave_the_nerd]

A 2Ghz i5, basically.

 

[Kalessian]

Maybe a stock FX-4300?

 

[Essence_of_War]

Yeah, I'd say the 2GHz i5 sounds about right.

http://www.anandtech.com/bench/product/53?vs=1261

If you adjust both scores to 4 GHz you get ~ 4630 compared to ~8707, so the haswell i5 is just under 2x faster.

I'd expect it to be within a few percent or so of the FX-4300 at stock, depending on whether or not the bench uses floats:
http://www.anandtech.com/bench/product/53?vs=700

 

[Abwx]

Compared to a current i5 it would be about 3.6 in 7zip and Fritzchess and 2.7 in Cinebench, obviously it would be close in integer code and distant in FP.

Edit : I m talking of CB R15, not the CB R10 wich can be found in AT benches numbers and wich is not optimised for CPUs newer than than the core 2.

 

[jhu]

hypothetically, what is a 4 ghz Q6600 comparable to these days?

According to my tests with custom compiled Povray, it's about equivalent to a 4 core, nonHT Haswell processor at 2.7 GHz (eg Core i5 4430S) or a 3 module AMD FX at 3.5 GHz (eg FX 6300).

 

[Yuriman]

From my tests, in gaming a stock Haswell i3 will (generally) be better due to stronger single-threaded performance, but the Core2 will be ahead in multithreaded tasks. I'd call moving to an i3 (from a 4ghz Core2) a sidegrade (on average) if you're not taking into account power savings and platform improvements.

 

[lavaheadache]

From my tests, in gaming a stock Haswell i3 will (generally) be better due to stronger single-threaded performance, but the Core2 will be ahead in multithreaded tasks. I'd call moving to an i3 (from a 4ghz Core2) a sidegrade (on average) if you're not taking into account power savings and platform improvements.

strictly performance speaking

 

[Roland00Address]

Who the hell has a 4 ghz core 2 q6600 (instead of 3.4 ghz or 3.6 ghz) and has not upgraded to anything newer? It takes serious cooling, lots of volts, and a lucky chip to hit 4ghz on a 65 nm core 2 quad. We are talking a 444 fsb with a locked 9x multiplier, or 500fsb with a 8x multiplier.

I call shenanigans.

 

[SlowSpyder]

Who the hell has a 4 ghz core 2 q6600 (instead of 3.4 ghz or 3.6 ghz) and has not upgraded to anything newer? It takes serious cooling, lots of volts, and a lucky chip to hit 4ghz on a 65 nm core 2 quad. We are talking a 444 fsb with a locked 9x multiplier, or 500fsb with a 8x multiplier.

I call shenanigans.

Until not very long ago I ran a 4GHz PhII (though a hexcore). It was probably only marginally faster than a 4GHz C2Q with the exception of software that could take advantage of the extra couple of cores I had.

 

[skipsneeky2]

Lava i would have a blast giving you a cinebench r15 benchoff with my gimped x4 9150e@1.8GHz.This is my back up cpu atm and its sitting in a very special 780g motherboard that basically locked its ht frequency to 2.0 or 1GHz instead of the usual 1.6GHz

I am pretty certain my 9150e is the slowest quad core chip ever produced ever.

 

[crashtech]

User "Conroe" cranked an E5450 up to 4.27GHz, and it's slightly ahead of a stock i5-4430 in Cinebench 11.5. Of course the Xeon feature 45nm but it's not a huge perf increase over 65nm C2Q:

http://forums.anandtech.com/showthread.php?t=2361017

 

[Makaveli]

Until not very long ago I ran a 4GHz PhII (though a hexcore). It was probably only marginally faster than a 4GHz C2Q with the exception of software that could take advantage of the extra couple of cores I had.

apples vs oranges friend.

The phenom II is newer than Core2quad.

Your talking 65nm vs 45nm.

That AMD chip hitting 4 Ghz isn't a big deal or hard.

A Q6600 being stable at 4 Ghz is a big deal and alot harder to do.

 

[SlowSpyder]

apples vs oranges friend.

The phenom II is newer than Core2quad.

Your talking 65nm vs 45nm.

That AMD chip hitting 4 Ghz isn't a big deal or hard.

A Q6600 being stable at 4 Ghz is a big deal and alot harder to do.

Sure, I understand that. Just talking about it from an absolute performance perspective. The PhII might be more likely to clock that high, but I think the IPC of the PhII was pretty close. So at similar clocks, I'd expect benches to be similar, even if power use might not be and / or the likelihood of achieving 4GHz clocks isn't the same.

 

[Makaveli]

Sure, I understand that. Just talking about it from an absolute performance perspective. The PhII might be more likely to clock that high, but I think the IPC of the PhII was pretty close. So at similar clocks, I'd expect benches to be similar, even if power use might not be and / or the likelihood of achieving 4GHz clocks isn't the same.

For absolute performance agreed.

IPC was close then skyrocketed after that in the distance future.

The load power the Q6600 would be pulling at that speed over the PhII should be noticeable.

What kind of HSF were you using on your setup?

 

[SlowSpyder]

For absolute performance agreed.

IPC was close then skyrocketed after that in the distance future.

The load power the Q6600 would be pulling at that speed over the PhII should be noticeable.

What kind of HSF were you using on your setup?

Scythe Ninja Copper. Nothing magical, quality cooler, but hardly the best even in its day. I admit, I bought it more because I thought it looked cool versus getting the best I could have gotten at the time.

 

[lavaheadache]

Who the hell has a 4 ghz core 2 q6600 (instead of 3.4 ghz or 3.6 ghz) and has not upgraded to anything newer? It takes serious cooling, lots of volts, and a lucky chip to hit 4ghz on a 65 nm core 2 quad. We are talking a 444 fsb with a locked 9x multiplier, or 500fsb with a 8x multiplier.

I call shenanigans.

my old one does, I'm just looking to put it back into commission as a back up. Though I am going to dial back the clock speed to 3.6. I was just trying to bring some excitement to the topic

 

[SPBHM]

Q9550 @ 4GHz here compared to CPUs as new as haswell
http://pclab.pl/art50000-54.html

Q6600 would be on average around 5% slower at this clock I think...

edit:
for what is worth, I just remembered I have saved results I got on CB11.5 with C2Q 65nm and sandy bridge with no HT at exactly 3.00GHz, i5 scored 4.76 points, the C2Q 3.34, 4GHz would have the same kind of difference I guess, that's just sandy bridge (and it can go at much higher clocks)...

 

[Roland00Address]

Sure, I understand that. Just talking about it from an absolute performance perspective. The PhII might be more likely to clock that high, but I think the IPC of the PhII was pretty close. So at similar clocks, I'd expect benches to be similar, even if power use might not be and / or the likelihood of achieving 4GHz clocks isn't the same.

I agree with that, my point is getting that type of clock speed would require a serious water loop (or more serious cooling), a person who knows what they are doing, and a lucky chip.

Now a Core 2 Duo E8400/8500/8600 or such on the other hand had a easier time reaching 4ghz but even that was hard.

Now just posting a 4 ghz Q6600 is different than a stable Q6600. If you can't have the chip stable for normal ever day use what is the point of comparing a 4ghz quad instead of something like a 3.6 ghz.

my old one does, I'm just looking to put it back into commission as a back up. Though I am going to dial back the clock speed to 3.6. I was just trying to bring some excitement to the topic

Sorry if I came off as rude. 3.6 ghz is often far more stable and far more reasonable. That extra 10% speed is often not worth the headache.

Yeah you did bring some excitement to the topic. I figured you asked for 4ghz thinking the numbers would be easier to calculate.

But like other people said a 2 ghz to 3 ghz core i5 would be similar in performance depending on the software used. Software that is written like crap and does not take advantage of 8 years of progress would perform closer to the high 2ghz core it (2.6, 2.8, 3.0). Better software would be 2 ghz. Things that take advantage of things like AVX we are talking an order of magnitude faster.

Or put another way. The cheap g3258 overclockable at stock speeds is much faster than the q6600. Sometimes at stock speeds 50% faster in multi thread, and even more so in single thread. The g3258 is also a simple chip to overclock. Of course the i3s, i5s, and i7s are better in practically everything.

 

[soccerballtux]

According to my tests with custom compiled Povray, it's about equivalent to a 4 core, nonHT Haswell processor at 2.7 GHz (eg Core i5 4430S) or a 3 module AMD FX at 3.5 GHz (eg FX 6300).

cool/interesting. that's a bit slower than I was expecting

 

[deanx0r]

I have a Q6600 @ 3.6GHz and recently picked up a Pentium G3258 clocked at 4.4GHz. Even with an older HD6970, there was significant improvements in games (smoother and steadier framerate). However, as others have mentioned, the Q6600 felt more responsive in multithreaded/multitasking scenarios. While the Q6600 served me well, it did so by dumping a significant amount of heat in the case.

I would call it a sidegrade and I am a bit disappointed that the overclocked G3258 cant even match a stock i3.

 

[tweakboy]

Get the Q6600 and save some money on purchase. gl

 

[RU482]

How does a 4ghz core 2 quad stack up these days?

I was gonna say "On top of an LGA 775 Socket", but I've got a strange sense of humor