New Mid-Priced Coolers from OCZ and Scythe: The Small and the X of It
by Wesley Fink on October 1, 2007 12:05 AM EST- Posted in
- Cases/Cooling/PSUs
Overclocking
As cooling solutions do a better job of keeping the CPU at a lower temperature, then it is reasonable to expect the overclocking capabilities of the CPU will increase. In each test of a cooler we measure the highest stable overclock of a standard X6800 processor under the following conditions:
CPU Multiplier: 11x (Stock 11x)
CPU Strap: 266 or 333
CPU voltage: 1.5875V at Max overclock
FSB Voltage: 1.40V
Memory is set to Auto timings on the P35. This removes memory as any kind of impediment to the maximum stable overclock. As FSB is raised the linked memory speed increases in proportion. For this reason memory speed is maintained as close as possible to CPU strap settings to prevent memory becoming an issue in CPU testing. This means 3.33GHz uses a 333 strap and runs 10x333, 3.73GHz runs a 333 strap and runs 11x339, and 3.90GHz is 333 strap and 354x11. The stock speed test is 266 strap and 11x266. We may move in the future to a 3.0GHz stock speed with a 333 strap and 9x333 settings for complete consistency, but since we are not measuring raw computer performance in our cooling benchmarking this becomes a moot point. The same processor is used in all cooling tests to ensure comparable results.
The OCZ Vendetta tops out at 3870MHz - near the top of tested air coolers. This matches the Thermalright Ultima-90 when it is running a 90mm fan, but it does not quite reach the 3.90 to 3.94GHz of the top coolers tested. However, for around $30 the OCZ Vendetta is able to match the best overclock that could be achieved with entry level water cooling with the Corsair Nautilus 500 or the Swiftech H2O 120 Compact. That is pretty impressive performance for a cooler aimed at midrange air. To be fair the operating temperatures across the speed range for the OCZ Vendetta are generally not as good as the water cooling kits, but the temps are still in the safe range for the X6800 CPU.
The Scythe Kama Cross is more like an Intel retail cooler on steroids (admittedly a light dose), reaching 3.80GHz with the stock 100mm fan and 3.83GHz with the 120mm fan. With the starting price of the Kama Cross actually being higher than the Vendetta, it is difficult to get excited about these overclocking results.
As stated many times, the overclocking abilities of the CPU will vary at the top, depending on the CPU. This particular CPU does higher FSB speeds than any X6800 we have tested, but the 3.90GHz top speed with the Thermalright Ultima-90 with 120mm fan is pretty average among the X6800 processors we have tested. A few of the other processors tested with the best air coolers reach just over 4GHz, but the range has been 3.8 to 4.0GHz. Stock cooling generally tops out 200 to 400 MHz lower, depending on the CPU, on the processors tested in our lab.
As cooling solutions do a better job of keeping the CPU at a lower temperature, then it is reasonable to expect the overclocking capabilities of the CPU will increase. In each test of a cooler we measure the highest stable overclock of a standard X6800 processor under the following conditions:
CPU Multiplier: 11x (Stock 11x)
CPU Strap: 266 or 333
CPU voltage: 1.5875V at Max overclock
FSB Voltage: 1.40V
Memory is set to Auto timings on the P35. This removes memory as any kind of impediment to the maximum stable overclock. As FSB is raised the linked memory speed increases in proportion. For this reason memory speed is maintained as close as possible to CPU strap settings to prevent memory becoming an issue in CPU testing. This means 3.33GHz uses a 333 strap and runs 10x333, 3.73GHz runs a 333 strap and runs 11x339, and 3.90GHz is 333 strap and 354x11. The stock speed test is 266 strap and 11x266. We may move in the future to a 3.0GHz stock speed with a 333 strap and 9x333 settings for complete consistency, but since we are not measuring raw computer performance in our cooling benchmarking this becomes a moot point. The same processor is used in all cooling tests to ensure comparable results.
The OCZ Vendetta tops out at 3870MHz - near the top of tested air coolers. This matches the Thermalright Ultima-90 when it is running a 90mm fan, but it does not quite reach the 3.90 to 3.94GHz of the top coolers tested. However, for around $30 the OCZ Vendetta is able to match the best overclock that could be achieved with entry level water cooling with the Corsair Nautilus 500 or the Swiftech H2O 120 Compact. That is pretty impressive performance for a cooler aimed at midrange air. To be fair the operating temperatures across the speed range for the OCZ Vendetta are generally not as good as the water cooling kits, but the temps are still in the safe range for the X6800 CPU.
The Scythe Kama Cross is more like an Intel retail cooler on steroids (admittedly a light dose), reaching 3.80GHz with the stock 100mm fan and 3.83GHz with the 120mm fan. With the starting price of the Kama Cross actually being higher than the Vendetta, it is difficult to get excited about these overclocking results.
As stated many times, the overclocking abilities of the CPU will vary at the top, depending on the CPU. This particular CPU does higher FSB speeds than any X6800 we have tested, but the 3.90GHz top speed with the Thermalright Ultima-90 with 120mm fan is pretty average among the X6800 processors we have tested. A few of the other processors tested with the best air coolers reach just over 4GHz, but the range has been 3.8 to 4.0GHz. Stock cooling generally tops out 200 to 400 MHz lower, depending on the CPU, on the processors tested in our lab.
Facebook
Twitter
RSS
22 Comments
View All Comments
Wesley Fink - Monday, October 1, 2007 - link
There is only one game available that may possibly perform better with a quad CPU. Our original thinking was that we would wait to move to quad testing until games are launched that give us a reason to buy a quad-core - somewhere in the future.However, the point is well-taken that quads do generate more heat than dual-core processors, so we will be doing a comparison in the next few weeks on a range of coolers tested on a quad-core CPU.
Acanthus - Monday, October 1, 2007 - link
That is fantastic, not all of us buy quads for gaming :DI have one for encoding.
strikeback03 - Tuesday, October 2, 2007 - link
same here, though it would be nice if Premiere would make use of more than one core.Spacecomber - Monday, October 1, 2007 - link
The graphs and the discussion of them seemed out of synch, especially on page 5. It's as if the chart being discussed didn't get included or the discussion is meant for another section.Anyway, I got confused at that point, just looked at the graphs, and drew my own conclusions from there on. ;-)
Wesley Fink - Monday, October 1, 2007 - link
The Intel Retail results on p.5 had not been updated for CoreTemp results. That has now been updated. Hopefully the commentary now matches the graphs.Spacecomber - Monday, October 1, 2007 - link
Yup, the previous bar for the intel heatsink said something like 41 deg, which left me scratching my head.The discussion of how the cooling scaled with higher clock speeds also seemed out of place on that page, since there is no graphical representation of that until you get to page 7. Perhaps you were just making those comments paranthetically on page 5, but I wasn't clear why it was being discussed where all the results being graphically presented were for default clock speeds.
FrankThoughts - Monday, October 1, 2007 - link
...that took one look at the Scythe "I'm an X! Isn't that AWESOME!" design and immediately figured performance was going to suck? Repeat after me: gimmicky cooler designs do NOT work well! Just look at the first image of the cooler: all the closely packed fins, lots of gap between the fan and the fins, and you can already guess that most of the air so going to go around the fins rather than through them.Maybe a plastic shroud around the HSF would have helped, but even then a large amount of air would just go through the center gap. This is one of those designs that looks nifty but has some bassackwards thermodynamic "theory" at its core.
strikeback03 - Tuesday, October 2, 2007 - link
I'm just surprised they let something out that sucks this bad. If the performance were at least a little better than stock some might buy it for the looks, but this performance is just embarrassing to Scythe.puffpio - Monday, October 1, 2007 - link
I agreeYou can immediately look at it and see that the air the fan blows is going to go AROUND the heat sink...basic fluid dynamics..path of least resistance
But also w/ the OCZ cooler, they cut out some heat sink fins to make a curve shape...so you loose thermal capacity and gain aesthetics?
Someone needs to design an enthusiast heat sink and fan that consulted a thermal and aerodynamic engineer...perhaps tapping into the skill set of people who design car radiators..
KazenoKoe - Wednesday, October 3, 2007 - link
I would like to know how the Scythe cooler performs with 2 smaller fans, one for each heatsink, instead of one fan at an odd angle.