Affordable Pelter Cooling

no such thing as affordable pelt cooling. not effective pelt cooling, anyhow.

that pelt is not enough to keep a CPU usefully cooler than air alone, let alone water, and that heatsink can't keep the pelt's hotside cool enough for optimal opperation without a unreasonably loud fan.

that doesn't even get into the power requirements, or the difficulty of getting the pelt's waste heat ouf of the case. also, whoever wrote that spin-piece should be shot for calling pelts efficient. pelts are energy guzzling little monsters that will try to kill both you and your system unless you are faster and meaner than they are. :) another thing: condensation from straight watercooling? please.......teach this arts major some physics.

might as well just get a nice heatsink, rather than getting that thing. swiftech tried the same kind of approach, and it just doesn't deliver.
 
Well, I purchased one on the eBay from them. It works great. It is very easy to install, just like a regular headsink with a controller. I was able to overclock my AMD 4000+ (ASUS A8N SLI Premium) by more than 15% without any problem. The temperature stays at about 105 F idle and 125 with full load. I was able to achieve over 12800 on 3DMark2005 at default setting! One thing I have notice is that it is regulated by some sort of algorithm, in which one the temperature hits 125 F, it drops 20 F in matter of seconds. It doesn't come cheap at $150 but I think it is worth it, because I am getting FX57 ($1100) performance out of my 4000+ ($350). Anyway, I was told that several reviews are due out soon. If you don't trust my result, you certainly can wait for one of the reviews. Good luck.
 
Lord Vicious said:
Well, I purchased one on the eBay from them. It works great. It is very easy to install, just like a regular headsink with a controller. I was able to overclock my AMD 4000+ (ASUS A8N SLI Premium) by more than 15% without any problem. The temperature stays at about 105 F idle and 125 with full load. I was able to achieve over 12800 on 3DMark2005 at default setting! One thing I have notice is that it is regulated by some sort of algorithm, in which one the temperature hits 125 F, it drops 20 F in matter of seconds. It doesn't come cheap at $150 but I think it is worth it, because I am getting FX57 ($1100) performance out of my 4000+ ($350). Anyway, I was told that several reviews are due out soon. If you don't trust my result, you certainly can wait for one of the reviews. Good luck.
Click to expand...
so, converting to real units, you're saying that it "works" because it is giving you IDLE of 40.5C and LOAD of 51.6C..............what the hell are you smoking? a good air cooler can do as well or better.
 
Lord Vicious said:
Well, I purchased one on the eBay from them. It works great. It is very easy to install, just like a regular headsink with a controller. I was able to overclock my AMD 4000+ (ASUS A8N SLI Premium) by more than 15% without any problem. The temperature stays at about 105 F idle and 125 with full load. I was able to achieve over 12800 on 3DMark2005 at default setting! One thing I have notice is that it is regulated by some sort of algorithm, in which one the temperature hits 125 F, it drops 20 F in matter of seconds. It doesn't come cheap at $150 but I think it is worth it, because I am getting FX57 ($1100) performance out of my 4000+ ($350). Anyway, I was told that several reviews are due out soon. If you don't trust my result, you certainly can wait for one of the reviews. Good luck.
Click to expand...

HI I just joined so I could pimp my product. Thanks!
 
I am not smoking anything. Notice that this is the temperature when the 4000+ is overclocked at 230MHz FSB. In addition, since the multiplier can't be changed, the entire system is actually running at 15% over stock. Can any of the air cooler do that? If you can find me an air cooler that can sustain a 15% overclock under full load over a period of 96 hours, I will get one right now.
 
oh dear... i think most air coolers worth their weight in copper can sustain a 15% overclock on their own. I think my friend is running the auto 15% oc on his asus board using the stock hsf with his a64.

hate to do this to someone new in the game but keep trying!
 
Check my email out. You will find I actually work for their competitor :) . I don't have a reason to pimp their product except sharing my experience with you.
 
Please show me the whole system configuration and the bios set-up of your friend's system. By the way, the highest auto overclocking setting on the Asus board is 10%. I don't know how you get 15%. Any how, maybe I am new at this, but I would love to learn how your friend does it.
 
Lord Vicious said:
I am not smoking anything. Notice that this is the temperature when the 4000+ is overclocked at 230MHz FSB. In addition, since the multiplier can't be changed, the entire system is actually running at 15% over stock. Can any of the air cooler do that? If you can find me an air cooler that can sustain a 15% overclock under full load over a period of 96 hours, I will get one right now.
Click to expand...
22% OC using stock air cooler on my 3000+ a64. 1.8 -> 2.2, while i have my real pelt cooler offline pending some bug fixes.

i idle at around 30, load gets up to 41.

a good heatsink like the thermalright XP-90 can do one heck of a lot better.

that pelt cooler is junk, plain and simple. i am sorry to say that you have wasted your money.
 
Lord Vicious said:
Please show me the whole system configuration and the bios set-up of your friend's system. By the way, the highest auto overclocking setting on the Asus board is 10%. I don't know how you get 15%. Any how, maybe I am new at this, but I would love to learn how your friend does it.
Click to expand...

You would find the 15% just as you would finding any other percentage. Use those algebra skills!
 
Bassist-X said:
You would find the 15% just as you would finding any other percentage. Use those algebra skills!
Click to expand...
i thnk that it's more that he doesn't know how to set up a bios......
 
Lord Vicious said:
I am not smoking anything. Notice that this is the temperature when the 4000+ is overclocked at 230MHz FSB. In addition, since the multiplier can't be changed, the entire system is actually running at 15% over stock. Can any of the air cooler do that? If you can find me an air cooler that can sustain a 15% overclock under full load over a period of 96 hours, I will get one right now.
Click to expand...

You my friend have been seriously duped. I have a very POOR overclocking chip...yet I can get a 20% OC out of mine on stock aircooling. My temps never go above 48C when overclocked. (2.0 -> 2.4)
 
DFI Daishi said:
so, converting to real units, you're saying that it "works" because it is giving you IDLE of 40.5C and LOAD of 51.6C..............what the hell are you smoking? a good air cooler can do as well or better.
Click to expand...
My Venice 3000+ is doing 250x9 with an idle of 32C and a load of 39-40C with an XP-90 that included a 92mm Thermalright branded fan. That's 25%.

I did a 10% OC at stock voltage with the stock HSF and my temps never got above 46 full-tilt.

pwnd!
 
Of course I know algebra. The guys said 15% profile in AUTO mode. In the auto-overclocking, the 10% profile is the highiest AUTO overclocking setting.
 
GilmourD said:
My Venice 3000+ is doing 250x9 with an idle of 32C and a load of 39-40C with an XP-90 that included a 92mm Thermalright branded fan. That's 25%.

I did a 10% OC at stock voltage with the stock HSF and my temps never got above 46 full-tilt.

pwnd!
Click to expand...

How about your DDR setting and HT setting? Are you telling me your HT setting is at 250 X 5 X 2 = 2500 MHz? By the way, your 3000+ Venice should run cooler anyway because the thermal power on it is only 67W. The 4000+ runs at 89W. Thus temperature difference.
 
DFI Daishi said:
22% OC using stock air cooler on my 3000+ a64. 1.8 -> 2.2, while i have my real pelt cooler offline pending some bug fixes.

i idle at around 30, load gets up to 41.

a good heatsink like the thermalright XP-90 can do one heck of a lot better.

that pelt cooler is junk, plain and simple. i am sorry to say that you have wasted your money.
Click to expand...

How about your DDR setting and HT setting? Are you telling me your HT setting is at 250 X 5 X 2 = 2500 MHz? By the way, your 3000+ Venice should run cooler anyway because the thermal power on it is only 67W. The 4000+ runs at 89W. Thus temperature difference.
 
DFI Daishi said:
22% OC using stock air cooler on my 3000+ a64. 1.8 -> 2.2, while i have my real pelt cooler offline pending some bug fixes.

i idle at around 30, load gets up to 41.

a good heatsink like the thermalright XP-90 can do one heck of a lot better.

that pelt cooler is junk, plain and simple. i am sorry to say that you have wasted your money.
Click to expand...

Can you run full-load for 72 hours without overheating at that OC? Stock air-cooling can not achieve the rapid cooling as active cooling like TEC or water cooling can do.
 
Couple of thoughts for you, Lord Vicious.

First, we generally don't post 5 times in a row here, that makes you look like a post whore. Try to consolidate whatever's on your mind to a post or two at a time.

Second, as someone who has acknowledged you don't know how to manually overclock, you shouldn't argue with DFI_Dashi. He's built his own custom pelt systems, and, like most of us on this board, can overclock manually.

If you want to see the average temps people are getting with their 4000's make a thread in the AMD section... you'll find your temps are merely average to a bit higher than normal.

Last, The cooling on your CPU has nothing to do with how well your RAM overclocks, and HTT settings don't mean much for performance.

See how I did that all in one post? ;)
 
Jonsey said:
Couple of thoughts for you, Lord Vicious.

First, we generally don't post 5 times in a row here, that makes you look like a post whore. Try to consolidate whatever's on your mind to a post or two at a time.

Second, as someone who has acknowledged you don't know how to manually overclock, you shouldn't argue with DFI_Dashi. He's built his own custom pelt systems, and, like most of us on this board, can overclock manually.

If you want to see the average temps people are getting with their 4000's make a thread in the AMD section... you'll find your temps are merely average to a bit higher than normal.

Last, The cooling on your CPU has nothing to do with how well your RAM overclocks, and HTT settings don't mean much for performance.

See how I did that all in one post? ;)
Click to expand...


You misunderstood me. I mean you can manually overclock more than 10% with that particular board, but not using the auto-overclocking function. I do know how to manually overclock my system. And it does make a big difference with the HT setting. See http://images.anandtech.com/reviews/cpu/amd/athlon64/veniceoverclocking/overclocksettings.png and you will understand what different HT setting can do to a system. The temperature of the CPU depends greatly on how the whole system is set-up. Simply changing your HT setting from X5 to X4 and you will see a great difference in your performance. Please read the anandtech review on the overclock and you will have a better understand.

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2548&p=11
 
Lord Vicious said:
How about your DDR setting and HT setting? Are you telling me your HT setting is at 250 X 5 X 2 = 2500 MHz? By the way, your 3000+ Venice should run cooler anyway because the thermal power on it is only 67W. The 4000+ runs at 89W. Thus temperature difference.
Click to expand...
running 1:1 with HTT set to 4x, as is standard when doing any substantial overclocking, locked multi on my core limits me to 9x, so i'm currently running 247 MHz.

as for how long i can maintain this overclock without overheating......i don't know how you define load, however i typically consider it to be at least an hour of prime95 torture test, and my temps don't go any higher than the aforementioned 41 C. regardless of how long i run, that's as hot as it gets. period. what would overheat, after 72 hrs. anyway? things get as hot as they are going to get after an hour of intensely hammering on them. running that long is typically testing for OC stability in terms of computataional accuracy, not as as function of trying to get hotter.

as for a pelt cooling "faster"..........that depends on a whole lot of different factors, however as a general rule, it is quite the opposite, resulting in very gradual temperature swings while the system is running, not fast ones. that cold plate doesn't change temperatures quickly.
 
Lord Vicious said:
You misunderstood me. I mean you can manually overclock more than 10% with that particular board, but not using the auto-overclocking function. I do know how to manually overclock my system. And it does make a big difference with the HT setting. See http://images.anandtech.com/reviews/cpu/amd/athlon64/veniceoverclocking/overclocksettings.png and you will understand what different HT setting can do to a system. The temperature of the CPU depends greatly on how the whole system is set-up. Simply changing your HT setting from X5 to X4 and you will see a great difference in your performance. Please read the anandtech review on the overclock and you will have a better understand.

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2548&p=11
Click to expand...


I'm not sure how that link proves anything. Afterall, they lowered their HT multi when they overclocked as well:

anandtech said:
We'll keep our HT bus speed at or below 1050MHz by adjusting the HT multiplier as we increase the CPU bus speed (and we may at times drop lower if that brings stability). We can use the 4X multiplier with up to a 260MHz bus, and 3X will get us up to a 350MHz CPU bus
Click to expand...
 
DFI Daishi said:
running 1:1 with HTT set to 4x, as is standard when doing any substantial overclocking, locked multi on my core limits me to 9x, so i'm currently running 247 MHz.

as for how long i can maintain this overclock without overheating......i don't know how you define load, however i typically consider it to be at least an hour of prime95 torture test, and my temps don't go any higher than the aforementioned 41 C. regardless of how long i run, that's as hot as it gets. period. what would overheat, after 72 hrs. anyway? things get as hot as they are going to get after an hour of intensely hammering on them. running that long is typically testing for OC stability in terms of computataional accuracy, not as as function of trying to get hotter.

as for a pelt cooling "faster"..........that depends on a whole lot of different factors, however as a general rule, it is quite the opposite, resulting in very gradual temperature swings while the system is running, not fast ones. that cold plate doesn't change temperatures quickly.
Click to expand...

The stock/default HT setting is 5, which gives the 2000MHz system bus. When you overclock so much for the CPU side, that is why you had to change to 4, which gives 250 X 4 X 2= 2000MHz. However, what I did was to overclock the whole system, which I had kept HT at 5 and resulting 230 X 5 X 2 = 2300MHz System bus. As I have explained before, you can and will see a big difference for the two settings. In addition, your Venice 3000+ have a thermal power of 67W as opposed to my 4000+ whihc have a thermal power of 89W. That makes the temperature of our CPU different.

As of the pelt cooling, this is what I see. There is a LCD control panel come with this cooler that shows the temperature of the CPU/System. As soon as the temperature reached to 125F, with in 10 seconds, you can see it drop below 105. So, that is why I said it changes temperature fast. This is my theory on the contolling unit: They designed with a set temperature range where they will activate pelt, and de-activate it as soon as it reaches either end of the pre-set point. By doing so, they can insure problem of condensation a thing of the past. That is just my take. I have no clue if this is what they did.
 
Jonsey said:
I'm not sure how that link proves anything. Afterall, they lowered their HT multi when they overclocked as well:
Click to expand...

yes. If you overclock like Daishi did, at 250MHz, then, the X 4 setting is cool because your system bus still runs at 2000MHz. However, if your overclocking is in between, say 222 to 245, then, the X 4 setting actually slows down your system bus lower than default. This in terms will lower your performance, hence reduce your intended overclocking performance.
 
Megadeth_Guy01 said:
His chip puts out 67W at the STOCK CLOCK SPEED. When overclocked that number rises considerably.
Click to expand...

Yes, and as I overclock my CPU, my number rises considerably too. With his overclocking, his CPU is running at 2.2GHz. My CPU is running at 2.8GHz after my overclocking. Now, do you see my point?
 
Lord Vicious said:
Yes, and as I overclock my CPU, my number rises considerably too. With his overclocking, his CPU is running at 2.2GHz. My CPU is running at 2.8GHz after my overclocking. Now, do you see my point?
Click to expand...

Yes, but that could have been done on air too, and been damn stable.
 
Lord Vicious said:
The stock/default HT setting is 5, which gives the 2000MHz system bus. When you overclock so much for the CPU side, that is why you had to change to 4, which gives 250 X 4 X 2= 2000MHz. However, what I did was to overclock the whole system, which I had kept HT at 5 and resulting 230 X 5 X 2 = 2300MHz System bus. As I have explained before, you can and will see a big difference for the two settings.
Click to expand...
i can also see a big difference between running at a 230 MHz FSB and running an almost 250 MHz. overclocking is about testing and modifying a whole suite of variables to net you the best performance in the applications that you use, games or otherwise. reducing my HTT to 4x and going for a higher FSB and core clock nets me better performance than HTT of 5x and a lower clock does.

Lord Vicious said:
In addition, your Venice 3000+ have a thermal power of 67W as opposed to my 4000+ whihc have a thermal power of 89W. That makes the temperature of our CPU different.
Click to expand...
you keep repeating this as if you think that it is a strong point in your favour. dude, i'm running at LOWER temperatures with you are, with a stock cooler. regardless of my CPU having a lower power output, your process or could be getting better temperatures than what you are getting now, with a good, aftermarket, air cooler.

Lord Vicious said:
As of the pelt cooling, this is what I see. There is a LCD control panel come with this cooler that shows the temperature of the CPU/System. As soon as the temperature reached to 125F, with in 10 seconds, you can see it drop below 105. So, that is why I said it changes temperature fast. This is my theory on the contolling unit: They designed with a set temperature range where they will activate pelt, and de-activate it as soon as it reaches either end of the pre-set point. By doing so, they can insure problem of condensation a thing of the past. That is just my take. I have no clue if this is what they did.
Click to expand...
you're conclusion isn't unreasonable, howerer i would like to point out two things: 1) that is not typical of a pelt config, since a pelt is usually running constantly. don't try to make a general comment about pelts cooling faster, because they usually don't, even if your specific situation is a bit different. 2) rapid temperature swings are bad. they result in a lot of mechanical stress on the core and the chip package. the swings that you are seeing are probably not big enough to pop the die off, but don't act like temps swings are a sign of quality and goodness.

you might reall want to browse around and see what kind of temps and OCs people are running with your proc on straight air. we are telling you that people out there are doing as well or better with air, and you are telling us that we don't understand and about how different our own situations are from yours. you don't want to believe what we are telling you. fine, go find out and do the legwork yourself as to what others around here are experiencing, and back up what you are saying.
 
Lord Vicious said:
yes. If you overclock like Daishi did, at 250MHz, then, the X 4 setting is cool because your system bus still runs at 2000MHz. However, if your overclocking is in between, say 222 to 245, then, the X 4 setting actually slows down your system bus lower than default. This in terms will lower your performance, hence reduce your intended overclocking performance.
Click to expand...
unless you are using your computer for something pretty unusual, the higher core speed will net you better real world performance gains.

it would be nice to run 5x HTT and 300 FSB, however since we have to tune our systems according to what the hardware can and cannot do.......stop talking about the dream theory of what yields better CPU performance, and get the best performance that your actual hardware can sustain via adjusting the sesttings as you have to. insisting on a 5x HTT limits your performance.
 
Megadeth_Guy01 said:
Yes, but that could have been done on air too, and been damn stable.
Click to expand...


It would be difficult to using only air for higher frequency CPU. That is why more and more water-cooling/liquid cooling solution is available now as opposed a few years ago. Even if the passive cooling can achieve the needed performance, the heatsink would be huge and the fan would be very noisy
 
Lord Vicious said:
It would be difficult to using only air for higher frequency CPU. That is why more and more water-cooling/liquid cooling solution is available now as opposed a few years ago. Even if the passive cooling can achieve the needed performance, the heatsink would be huge and the fan would be very noisy
Click to expand...

Passive cooling = no fan FYI.

And no you are wrong again there, people are running Thermalright XP/SI-120's and 90's with pretty quiet fans yielding overclocks the same and higher than yours. You've been ripped off, and don't want to accept it. Typical.
 
If you want to go by percent increases, eclipse's opteron 144 overclocked 50% on stock cooling.

edit: and it didn't even magically blow up after 2 days, like you claim air cooling does
 
DFI Daishi said:
unless you are using your computer for something pretty unusual, the higher core speed will net you better real world performance gains.

it would be nice to run 5x HTT and 300 FSB, however since we have to tune our systems according to what the hardware can and cannot do.......stop talking about the dream theory of what yields better CPU performance, and get the best performance that your actual hardware can sustain via adjusting the sesttings as you have to. insisting on a 5x HTT limits your performance.
Click to expand...

I agree with you too, as I have found that this is not a normal pelt. The controlling unit actually changes current to the pelt to optimize power and temperature. I also agree to you that you can only do what your hardware allows you to. I am not here to fight with all of you, but just to share what I have. I would like to see someone with a overclocked 4000+, as I have read many places that as you reach higher on the Athlon CPU, it is harder to overclock. I also have heard that FX 57 is not so good as an overclock CPU.
 
Without looking at it from a cooling point of view and more from deductive reasoning, if this TEC is so effective, why is it that the only place you can find it sold is on ebay?

I'm guessing you bought it from this fellow: http://cgi.ebay.com/Vigor-AMD-Athlo...815924920QQcategoryZ80150QQrdZ1QQcmdZViewItem

Shady ebay dealer said:
With water-cooling solutions, many experts have raised concerns over disadvantages, such as possible water leakage, condensation and its efficiency dependence on room temperature.
Click to expand...
Neat, I didn't know WC systems can have condensation....And wouldn't contradict his next statement of its dependence on room temperature?

You shouldn't go by what that display tells you, you should check your CPU sensor temp...For all you know, it could just be a a HSF with a display that shows numbers ranging from 104-125F...
 
As an eBay Associate, HardForum may earn from qualifying purchases.
Megadeth_Guy01 said:
Passive cooling = no fan FYI.

And no you are wrong again there, people are running Thermalright XP/SI-120's and 90's with pretty quiet fans yielding overclocks the same and higher than yours. You've been ripped off, and don't want to accept it. Typical.
Click to expand...

Not here to argue with you or playing with words. Please send me a link of a 4000+ CPU overclocked to FX-57 performance using a stock fan. I would be happy. By the way, this TEC is being air-cooled too. I have no problem with air-cooling. I just want to see a real life example.
 
CoW]8(0) said:
Without looking at it from a cooling point of view and more from deductive reasoning, if this TEC is so effective, why is it that the only place you can find it sold is on ebay?

I'm guessing you bought it from this fellow: http://cgi.ebay.com/Vigor-AMD-Athlo...815924920QQcategoryZ80150QQrdZ1QQcmdZViewItem


Neat, I didn't know WC systems can have condensation....And wouldn't contradict his next statement of its dependence on room temperature?

You shouldn't go by what that display tells you, you should check your CPU sensor temp...For all you know, it could just be a a HSF with a display that shows numbers ranging from 104-125F...
Click to expand...

I agree with you on the display. I should check on that with other means. As far as the only place to buy this, have anyone of you heard of Thermaltake's Subzero 4G? That is something similar.
 
As an eBay Associate, HardForum may earn from qualifying purchases.
You must log in or register to reply here.
Back
Top