LoL, I can give you a "Cherrypicked" i7 that won't even go above 3.2ghz.
Srsly, Asus saying that is just plain idiocy. You can't get EVERY i7 up to 4ghz.
If they would send me a board with that claim, I'd ask for a new one untill they actually gave me a processor to go with it for the same price. Hey, it's their claim :p
i thought people were claiming 920 D0 were easier to overclock. seems like those are false claims. my 860 does 4Ghz at 1.3v, 3.8Ghz at 1.23v.
although i must say, the Corsair H50 isn't that good. at first it's temperature are amazing, but after you load the CPU for 2 hours, you'd see the temperature slowly going up. my 4Ghz has load temperature of 75c initially, but it slowly goes up and tops after 2 hours at 85c. i expect it's because the fans are not powerful enough to remove the heat or the radiator is not big enough.
however, despite it going over 80c when using Linx (Lin pack algorithm by Intel) i've never had over 60c in every day uses. i know because i've got Everest logging temperature into CSV.
Originally Posted by barndoor101 surely you should start low and creep higher instead of the other way round?
Effect of time - trying to nail something that will definitely do the job within the time I have available. In an ideal world, yes, we'd take the long winded iterative approach.
although i must say, the Corsair H50 isn't that good. at first it's temperature are amazing, but after you load the CPU for 2 hours, you'd see the temperature slowly going up. my 4Ghz has load temperature of 75c initially, but it slowly goes up and tops after 2 hours at 85c. i expect it's because the fans are not powerful enough to remove the heat or the radiator is not big enough.
The HD50s cooling ability is VERY much dependant on what case it is put in.
For example I originally had it a Jeantech Phong which had the normal push pull confirguratiion. Thanks to the H50s intake requirement I simply had to change cases. Now my system is housed with a Fractal Design R2 (which has two 120mm intakes, 2 x 120mm exhausts and the H50 acting as an intake) my temps have dropped considerably.
Of course the i7 CPUs put out so much heat (especially once overclocked) that temperates will still be high (relatively speaking) even with a high end water cooling setup. The H50 is better than most high end air coolers (especially when you have higher than average ambient temperatures) but it can't perform miracles.
tbh i rarely take the long winded approach myself, if you know the chip and the board you can usually get around about the right voltage and tweak downwards. nothing will explode, trust me.
I've currently got my P6TD Deluxe and 920 D0 at 4GHz on water with:
* vCore: 1.225V
* CPU Uncore: Auto
* DIMM: Auto
* PLL: Auto
* IOH: Auto
* ICH: Auto
* CPU Clock Amplitude: Auto
* Base Clock: 191MHz
* CPU Multiplier: 21x
* Memory Clock: 1,904MHz (6GB G.Skill Trident PC3-16000C9)
* Loadline Calibration: Enabled
* Memory Timings: Auto
Speedstep and Spread Spectrum options disabled. HT enabled.
Max temp after 6 hours Prime95 = 55ºC (ambient 20ºC).
I bought this board after reading the CPC review, well pleased with it. I've only been playing with it for a few days and I'm expecting to get a fair bit higher than that when I've figured out what everything does. I'm starting low and creeping higher.
Originally Posted by Big Elf I think I've been lucky getting a decent 920.
I think thats the main part of it. At home i have my 920 at 4ghz on a 1.3v vcore, whereas the 920 i have at work needs 1.35v to hit 3.6ghz (I wont take it any further as i dont wanna have to pay for it if it melts).
both are D0 steppings btw, but i guess i just got lucky with mine having a lower VID to start with.
Originally Posted by biebiep LoL, I can give you a "Cherrypicked" i7 that won't even go above 3.2ghz.
Srsly, Asus saying that is just plain idiocy. You can't get EVERY i7 up to 4ghz.
If they would send me a board with that claim, I'd ask for a new one untill they actually gave me a processor to go with it for the same price. Hey, it's their claim :p
I can't quote what Asus did say, but from my understaning they only said the board could do 4GHz stable, not "every i7".
My Gigabyte GA-EX58-UDR3 does 4.0 GHz stable all day no problem, and while I don't know if it will go for years and years this way, it's been a few months with no signs of trouble (unless I tweaked things). Turn off hyper threading and turbo mode, and be sure to keep on the power management settings so it doesn't run at full speed the whole time. I rarely go far past 70 degrees, and although it might not survive a week-long torture test, nobody runs their CPU that fast, that hot, for such a long period of time. Good luck getting a Core i7 to 100% capacity during regular use, for longer than a few seconds at once, anyway. If it runs stable with all programs and all regular activity, frankly I don't care too much if Prime95 crashes it after a few days.
If it doesn't run at 4ghz 24/7 what's the point in overclocking?
It's fast enough for general gaming at stock speeds. ( run mine at 3.4ghz for gaming- never seen an fps issue ever full HD resolution to HDTV ) ( 3.4 ghz is a tiny overclock I know but it does the job)
Only reason I overclock is to run programs that require 24/7 stability that actually benifit hugely
I read with interest your latest review of the Asus P6TD Deluxe. You mention in the review that the board now features a 'True 16+2-phase design when in fact or three versions of the board feature the 'True 16+2-phase design. As can be seen on Asus own website link: http://www.asus.com/compare.aspx
What was also interesting was the original Asus P6T Deluxe was faster than the V2 & P6TD with an overall media benchmarks score of 2,294 verses the V2, 2,170 & P6TD 2,236 of which you would expect the latest update of the board would have the best performance improvement? In the article you benchmarked the P6TD at 4.2GHz, a 21x multiplier & a 205MHz QPI. Should that not have given you a CPU speed of just over 4.3GHz? Where's the Turbo Key, which provide one-button overclocking which is also mentioned in the article?
They shipped the thing to you telling you it would get this CPU to 4GHz. They must reasonably have known this could not possibly be true. I don't think they need bit-tech to make their excuses for them.
i was looking for a motherboard and seeking bit-tech for reviews shortly before the review was up. Then i found out that here is a new version of this nice asus board i love so much, so i think i'll order one today
Yea I use the p6t7ws supercomputer which is basically the same board and it takes >1min to boot to SAS in raid 0. Most of the time is spent initializing the raid or something. (the part where it says spinning up disks) Are you saying that if I did not boot from the SAS it would be faster!? (assuming I still keep the raid 0)
Originally Posted by Jenny_Y8S RE: "silly SAS controller that added a minute to boot time for no good reason"
My boot time is <30 seconds with Win7 and I'm using the SAS controller for two of my storage drives.
Does your +1min to boot only apply when botting off a SAS drive?
Probably not. This is a common problem with motherboards that have an extra drive/raid controller chip that is in addition to what is provided by the chipset. Seems that during the boot this extra drvie/raid controller chip takes a while to go through its boot process if there are no drives connected to it. Seems to just keep searching for the drives for a pre set amount of time.
Comments 1 to 25 of 33
ReplySrsly, Asus saying that is just plain idiocy. You can't get EVERY i7 up to 4ghz.
If they would send me a board with that claim, I'd ask for a new one untill they actually gave me a processor to go with it for the same price. Hey, it's their claim :p
"Lie" is a misinterpretation. Please read the conclusion.
i thought people were claiming 920 D0 were easier to overclock. seems like those are false claims. my 860 does 4Ghz at 1.3v, 3.8Ghz at 1.23v.
although i must say, the Corsair H50 isn't that good. at first it's temperature are amazing, but after you load the CPU for 2 hours, you'd see the temperature slowly going up. my 4Ghz has load temperature of 75c initially, but it slowly goes up and tops after 2 hours at 85c. i expect it's because the fans are not powerful enough to remove the heat or the radiator is not big enough.
however, despite it going over 80c when using Linx (Lin pack algorithm by Intel) i've never had over 60c in every day uses. i know because i've got Everest logging temperature into CSV.
No that's what I started with. Just nailed the voltage like usual and went "ooops too high"
Effect of time - trying to nail something that will definitely do the job within the time I have available. In an ideal world, yes, we'd take the long winded iterative approach.
The HD50s cooling ability is VERY much dependant on what case it is put in.
For example I originally had it a Jeantech Phong which had the normal push pull confirguratiion. Thanks to the H50s intake requirement I simply had to change cases. Now my system is housed with a Fractal Design R2 (which has two 120mm intakes, 2 x 120mm exhausts and the H50 acting as an intake) my temps have dropped considerably.
Of course the i7 CPUs put out so much heat (especially once overclocked) that temperates will still be high (relatively speaking) even with a high end water cooling setup. The H50 is better than most high end air coolers (especially when you have higher than average ambient temperatures) but it can't perform miracles.
* vCore: 1.225V
* CPU Uncore: Auto
* DIMM: Auto
* PLL: Auto
* IOH: Auto
* ICH: Auto
* CPU Clock Amplitude: Auto
* Base Clock: 191MHz
* CPU Multiplier: 21x
* Memory Clock: 1,904MHz (6GB G.Skill Trident PC3-16000C9)
* Loadline Calibration: Enabled
* Memory Timings: Auto
Speedstep and Spread Spectrum options disabled. HT enabled.
Max temp after 6 hours Prime95 = 55ºC (ambient 20ºC).
I bought this board after reading the CPC review, well pleased with it. I've only been playing with it for a few days and I'm expecting to get a fair bit higher than that when I've figured out what everything does. I'm starting low and creeping higher.
I think I've been lucky getting a decent 920.
I think thats the main part of it. At home i have my 920 at 4ghz on a 1.3v vcore, whereas the 920 i have at work needs 1.35v to hit 3.6ghz (I wont take it any further as i dont wanna have to pay for it if it melts).
both are D0 steppings btw, but i guess i just got lucky with mine having a lower VID to start with.
i have overclocked enough of them to know -
its especially funny when Asus ask me for overclock settings to their boards
this is not news
a £90 gigabyte ud3r can do 4ghz
I can't quote what Asus did say, but from my understaning they only said the board could do 4GHz stable, not "every i7".
It's fast enough for general gaming at stock speeds. ( run mine at 3.4ghz for gaming- never seen an fps issue ever full HD resolution to HDTV ) ( 3.4 ghz is a tiny overclock I know but it does the job)
Only reason I overclock is to run programs that require 24/7 stability that actually benifit hugely
The v core is quiet high for what it is.
What was also interesting was the original Asus P6T Deluxe was faster than the V2 & P6TD with an overall media benchmarks score of 2,294 verses the V2, 2,170 & P6TD 2,236 of which you would expect the latest update of the board would have the best performance improvement? In the article you benchmarked the P6TD at 4.2GHz, a 21x multiplier & a 205MHz QPI. Should that not have given you a CPU speed of just over 4.3GHz? Where's the Turbo Key, which provide one-button overclocking which is also mentioned in the article?
Maybe you should try this again with a proper i7 920 D0?
No it bloody well isn't.
They shipped the thing to you telling you it would get this CPU to 4GHz. They must reasonably have known this could not possibly be true. I don't think they need bit-tech to make their excuses for them.
My boot time is <30 seconds with Win7 and I'm using the SAS controller for two of my storage drives.
Does your +1min to boot only apply when botting off a SAS drive?
Probably not. This is a common problem with motherboards that have an extra drive/raid controller chip that is in addition to what is provided by the chipset. Seems that during the boot this extra drvie/raid controller chip takes a while to go through its boot process if there are no drives connected to it. Seems to just keep searching for the drives for a pre set amount of time.
-
« Previous
-
1
-
2
-
Next »
Discuss in the forums