Most readers know that Intel's Core 2 Quad processors can beat its Core 2 Duos in a variety of multi-threaded applications, but they have less overclocking potential. How do the two technologies compare when both are taken to their clock limits?
- I would have liked to see the G0 Q6600 (priced within range of mere mortals) vs. the (similarly priced and same 9x multi) E6850
- There should have been some mention (and testing) of temperature differences.
...clock-for-clock, the quad is going to run hotter, potentially requiring a higher-dollar investment in cooling solution, so that needs to be factored in to the average user's buying decision as well (ie. the folks out here in the real world who don't get free high end water cooling setups).
Message edited by Spanki on 08-08-2007 at 06:05:28 PM
Lucky bugger, getting over 400Mhz fsb with the QX6850. Got my QX6850 on a Striker Extreme and I'm lucking if I can get the fsb stable over 350Mhz. I had to push my multiplier up to 11 to get near 4Ghz.
Yeah, it must be a huge difference in gaming experience having 310fps instead of 241 in F.E.A.R??? I can not understand why they didn´t test the gaming applications at least on 1600x1200. Who play games on 1024x768 with a Gf 8800gtx? Hands up?
Message edited by funnyvlad on 08-08-2007 at 06:56:01 PM
Most readers know that Intel's Core 2 Quad processors can beat its Core 2 Duos in a variety of multi-threaded applications, but they have less overclocking potential. How do the two technologies compare when both are taken to their clock limits?
They talk a lot about "stable" overclocks, but I wonder, just how hot are these suckers running? They never actually show the core temps for their quads. lol they're probably around 80-90 degrees C. Whether it is stable or not, it wont be forever at that temperature. And if any little thing goes wrong with your setup, there is no tolerance for it and you could be out an awful lot of money. Like that bridge in minneapolis...
Seriously, you people thinking about buying a Q6600 to overclock it, you better understand that you will be pushing some seriously high temperatures inside the cores, even at stock speeds. Under full load, my cpu temp sensor reads only 49C, but my core temps are up over 68C, at stock speeds. At 3GHz under full load, my cpu temp sensor reads 54C but my core temps are up around 85C. That's scary hot! I can't imagine it lasting very long at those temps. But what should be of greater concern is the sheer number of times the CPU will be forced to run that hot, and then relatively cooler when idling, then super hot again, then cool, etc. All I know is I'm not using my cpu to find out how many times it can take that before it breaks. That's what Tom's hardware is for! They need to get on it!
They talk a lot about "stable" overclocks, but I wonder, just how hot are these suckers running? They never actually show the core temps for their quads.
Upper 20's to low 30's, according to BIOS.
Message edited by Crashman on 08-08-2007 at 10:24:50 PM
Even with about $300 worth of custom watercooling parts with my watercooling system I dare not go above 3.8Ghz with my QX6850 as I'm already tapping 70C on the hottest core. (temps from Core Temp)
P.S. Anyone know the reason why the Core 2 Quads might have Core Temp temperatures separated almost exactly 3C from each other, such as core 0 = 40C, core 1 = 43C, core 2 = 46C, core 3 = 49C, under both all-cores idle and all-cores loaded situations?
I guess this may be a little beyond the scope of the article, but perhaps the test could be repeated using a different type of cooling technology. Swiftech makes great gear, no question, but I'm thinking a Mach II GT phase change cooler could yield better CPU overclocks.
Also, a G0 Q6600 needs to be included in the test.
---------------
Owner of the Dual Core Goliath.
Now building the Quad Core Colossus...
THG would have used an E6600 G0 stepping had it been available a few weeks back when the article was planned. Back then E6600 G0's were just starting to show up and nobody would guarantee what you'd get, not eve Intel.
Wait, are you kidding me? You get a top end C2D and C2Q, max overclock them with watercooling, and the games you choose to benchmark are F.E.A.R, CoD2, and Quake4? I've heard the same arguement over and over again with dual vs. quad for gaming and you made the same mistake as everyone else. You are using top end parts for 2-3 year old games.
I'm sure the gamers split between a q6600 and e6750 aren't thinking about if they can reach 300 FPS in CoD2. They are looking to the future with games that are *supposedly* going to run faster and with more features using a quad core over dual core. Granted it is impossible to benchmark a Half-Life2 ep:2 or Crysis until the fall, but you could have at least used games like Supreme Commander, Call of Juarez, and S.T.A.L.K.E.R. which are rumored to run better on a quad core.
Personally I am going with a q6600 g0 over any of the dual core. I find it dissapointing that I seem to remember reading interviews on Tom's Hardware with game developers talking about the benefits of quad core in so many different aspects other than FPS. Yet you guys are still chanting about dual core over quad core for gaming. Seriously, is anyone going to be able to tell 150FPS vs. 300 FPS? You prob could have got near the same performance from an overclocked e4400 with those games. If you're going to use bleeding edge PC parts, what exactly is the reason for not using bleeding edge games? How about an updated article with some better gaming comparisons please.
Message edited by Link649 on 08-09-2007 at 03:15:06 AM
Me I am waiting for that 8 core cpu. Hmmmm if adding two cores gives you an extra 38% improvement overall then an extra 4 cores should give you an extra 76% increase over a quad core. Quad core.......ha........you will be history soon enough...........
you are absolutly right Link649
and also how about articles with air cooled, which are less expensive than wattercooling but with great results ( thinking about the termalright ultra 120 extreme )
