Jump to content

vrodforums

Members
  • Content Count

    72
  • Joined

  • Last visited

Posts posted by vrodforums


  1. I must be lucky, I have been dealing with NewEgg for as long as I can remember. Never a problem at all and I even buy a lot of refurbished items (returns that are tested I'm sure). I just got my second x1900XTX for 469.00 and it's the best one I have seen. Overclocks like crazy and even though advertised that it was the card only it came in a box with all the cables and disc. I can only guess why someone would have returned it. :)

     

    I wish you luck in getting the issue resolved.


  2. This is almost funny, you guys put way too much faith in the on board sensors. I have tried the same cpu in three different DFI boards and two asus boards and the same cpu with the same cooler and it read different in every single board. One of the DFI boards read as much as 15C higher but that board now has my FX57 running at 3.05 24/7 and is rock solid with an air cooler showing 36C at idle and 50C loaded. Oh and all the boards tested read the exact same temp on the heat sink using a digital pyrometer.


  3. I found after I tried everything that my FX-60 likes the 13 multiplier as good as anything. It is rock solid at default voltage at 13x220 and lets me run my Corsair XMS PC3500 at 2-3-2-6. If I want to do some high benches I can throw in some Patriot PC4800 and it has no problem booting at 10x315 at all but the top of the stability for prime on air is 3.03 but requires 1.525 volts for that so I stick with the 13x220 @ default for 24/7 use.


  4. To be honest unless you plan to do multi tasking your going to find the single core is still IMHO the best route to go. I have both an FX-60 and an FX-57 and for all of the games my kids play (COD2, Fear, etc) the FX-57 matched with an x1900xt still offers the most bang for the buck. I also have several single core 146 and 148's that will all hit 2.8 @ 1.45 volts or less while the three 175 opterons I purchased not a one would push above 2.7 stable without pushing the voltage up above the 1.6 volt. There is a lot less luck required to get a good single core to overclock versus a good dual core.


  5. I have an x1900xt setup with an fx-57 and 4 HDD's on an ultra D using a 500w antec truepower and it is stable as a rock and my kids run it for not just hours but days at a time running at 3.05ghz and the card default with COD2 maxed out on the optima 100" in the living room. Never misses a beat or gives any problems. I have run it with several different bios and never had any issues. As a matter of fact I'm about to dump my SLi 7800GT's for another 1900 and might get the xtx but the card is the best gaming card I have ever seen. While the SLi's will score 1500 points higher in 3dmark a single 1900 will put the sli's to shame when you turn the graphics up to maximum.


  6. Well I'm not too much into Vapochills anyway, so change vapo with water :D I read the ones in XS cold bug at -58c (from s73ph3n and Pedro Rocha). Is a vapochill capable of so low temps? I thought you need cascade to get so low.

     

    I wasn't aware that cold bug had a defined temperature but I personally know people that have problems getting opterons to post with vapo chill. As far as water cooling goes I was making my own water blocks on a taig end mill in my garage a couple of years before I even heard of danger den and using everything from condensers to you name it for transfer and cooling. Until you start cooling the water down below ambient temp your overclock is going to be very limited.


  7. As always with OC nothing is assured. But you'll get more chances of getting 3.1ghz with 170+vapo than 3 with fx-60 and air.

     

    What if he purchased a 170 with a cold bug? I'm sorry I don't see vapo as being a 24/7 dependable unit for any period of time still, just too many things to factor into a unit like that.


  8. Opteron 146 CABYE 0540 FPMW

     

    DFI Lanparty UT nF4 Ultra-D (bios 11/14/2005)

    Opteron 146 10x290 2.9 @ 1.525

    2x512 1GB Patriot PC4800

    Nvidia 7800GT Evga 470x1100

    4x100GB SATA Maxtor 8mb cache raid 0

    Antec Truepower 500w

     

     

    Genie BIOS Settings:

    FSB Bus Frequency - 290

    LDT/FSB Frequency Ratio - AUTO

    CPU/FSB Frequency Ratio - 10

    PCI eXpress Frequency - 100Mhz

     

    CPU VID StartUp Value - StartUp

     

    CPU VID Control - 1.525v

    CPU VID Special Control - Auto

    LDT Voltage Control - 1.20v

    Chip Set Voltage Control - 1.50v

    DRAM Voltage Control - 2.70v

     

    DRAM Configuration Settings:

    DRAM Frequency Set - 200=RAM/FSB:01/01

    Command Per Clock (CPC) - Enable

    CAS Latency Control (Tcl) - 2.5

    RAS# to CAS# delay (Trcd) - 04 Bus Clocks

    Min RAS# active time (Tras) - 08 Bus Clocks

    Row precharge time (Trp) - 04 Bus Clocks

    Row Cycle time (Trc) - 16 Bus Clocks

    Row refresh cyc time (Trfc) - 16 Bus Clocks

    Row to Row delay (Trrd) - 02 Bus Clocks

    Write recovery time (Twr) - 02 Bus Clocks

    Write to Read delay (Twtr) - 02 Bus Clocks

    Read to Write delay (Trwt) - 03 Bus Clocks

    Refresh Period (Tref) - Auto

    Write CAS Latency (Twcl) - Auto

    DRAM Bank Interleave - Enabled

     

    DQS Skew Control - Auto

    DQS Skew Value - 0

    DRAM Drive Strength - Auto

    DRAM Data Drive Strength - Auto

    Max Async Latency - Auto

    DRAM Response Time - Fast

    Read Preamble Time - Auto

    IdleCycle Limit - 256 Cycles

    Dynamic Counter - Disable

    R/W Queue Bypass - 16 x

    Bypass Max - 04 x

    32 Byte Granularity - Disable(4 Bursts)

     

     

    ==========

     

     

    Prime95 + 3dMark2001SE + CPU-Z 1.31 + Everest 2.20:

     

    2900-146tmb.jpg

     

     

     

    3dMark2003:

     

    2900-146-03tmb.jpg

     

     

    3dMark2005:

     

    2900-146-05tmb.jpg

     

     

     

    ==========

     

     

     

    short description:

    My daughters machine, or was I gave her my FX-57 machine

    About average week 0540 chip

    I have a CAb2e 148 that is almost identical in performance

    DFI Lanparty UT nF4 Ultra-D (bios 11/14/2005)

    Opteron 146 10x290 2.9 @ 1.525

    2x512 1GB Patriot PC4800

    Nvidia 7800GT Evga 470x1100

    4x100GB SATA Maxtor 8mb cache raid 0

    Antec Truepower 500w

     


  9. How accurate is speedfan/MBM in comparison to proper methods.

     

    MBM is just using the sensors that the bios and other software uses. Run prime for 30 minutes and then stop it, does your temp go back to idle temp in less than 2 seconds? If you answered yes then I can assure you that your cpu temperature is not being monitored correctly. It would be impossible to take a mass of copper or aluminum and drop the temp that fast, even if the load is removed it would take at least two minutes to remove the energy that has been put into the mass. I know someone is going to tell me about physics and how they have a 120mm fan.

     

    I tried to post the formula for determining the time duration for air cooling to effect the mass but the code does not post correctly so you can view this link.

     

    http://www.aavidthermalloy.com/technical/p...mentation.shtml


  10. what exactly?

     

    Ive now got it @ default core and volts

     

    And 38oC idle 50oC load Prime blend?

     

    Jack I use a thermal laser to check mine but you can use a host of types of thermometers that use either thermal couples or hard probes that are not that expensive to verify temps. With three DFI boards not a one of them reads correct when checked with good temperature measuring tools.

     

    Hat I think you might be a little quick to pass judgment and I disagree with much of what you say for the record. You appear to be basing your information on bios or software readings and they simply are not correct with DFI boards. Great overclockers but terrible at reading temps.


  11. While you will read about those with the 3.0ghz dual cores not many of those really run on air cooling. Probably even less run 24/7 at those speeds but I'm sure a few really do. Do a search for the results of 4800+ overclocks and most seem to peak at around 2.8 for 24/7 operation with 1.4-1.5 volts applied. I have really only seen a couple that hit 3.0 and usually require more like 1.6 volts which isn't good on air cooling since the heat from the big dual cores is pretty intense. There is however nothing cast in stone and each chip seems to be a little different so your results might be better or worse. The only way to find out is to spend a little time reading the stickie guides and then see what yours will do. I run my machines 24/7 and usually will do some bench test higher but rarely do I apply extra voltage and settle for an overclock that runs at default voltage.


  12. vrodforums by "

     

    So your telling me that my fluke 87, my fluke 23 and my digital laser that all agree on the same temp are all incorrect and that my motherboard that cost less than any of the three of them is correct"

     

    do you mean an infrared laser thermometer? if you do those are not accurate aganist metal or shiny surfaces.........i know because i own 2 of these, if you read it should say so in the manual......which would explain the discrepancy

     

    while they aren't perfect, the temp sensors should be ok unless you really f'ed up your mobo.....i can't imagne a 54 degree difference........i honestly doubt that as accurate at all.....

     

    Not sure which model you have but mine is made just for metal and is dead on when double checked with a hard probe. While I agree that they are not off by 54 degrees they are off by as much as 15c (or atleast two of the boards I have are)and that is a big difference when your talking about stability of a cpu. I just read all these reports of idle temps being almost room temperature and I hate to break the news to people that a cpu on air cooling simply won't run at room temperature. So if your seeing 50C and you don't know how accurate your board really reads then you might just be hitting 65c or higher. Some report they hit a wall and won't go higher even with more volts but it may be they are hitting the thermal limits of the chip and not even know it.

     

    I currently have 4 different brands of 939 motherboards and the DFI boards read as much as 15c less than any of the other boards. If you want the best board for temp sensors I have one A8N32 that is dead on all the time when compared to a real measurement. I just can't believe that people think these chips heat up and cool down that quick to put any faith at all in these readings.


  13. So your going to tell me that 1.5 pounds of aluminum (since I run all copper) is going to cool down 10-15C in less then five seconds at ambient air temp because you have a 120mm fan on it? :) How many others notice the instant drop in temps when loads are removed using the motherboards sensor? How many of you notice the instant climb in temps by as much as 10-15C when a load is applied? Remember it also takes time to heat the mass of aluminum or copper as well. How many of you really think that is a good reading? :)


  14. @vrodforums With all due respect, man...

    bull crap!!! It's straight individual for each rig/cpu/cooler. Depends exclusively on your cooler and airflow inside the case. And, yes - MBM reports perfectly on any DFI NF4 board I've tested, this means at least 50 mobos. The difference between sensors and reality is no more than a degree. I can show screenies of my old [email protected] and the [email protected] with the Sonic Tower. The P4 couldn't drop to idle temps for about 10 mins w/ a 120mm fan @ 1000rpm. And the 144UP does it within a minute ! With a 120mm fan @ 560rpm... period. So - it's individual!

     

    So your telling me that my fluke 87, my fluke 23 and my digital laser that all agree on the same temp are all incorrect and that my motherboard that cost less than any of the three of them is correct (keep in mind I have three DFI boards and one of them the temps are different from the other two by over 10C using the exact same cpu and cooler on the bench without a case at all) ? I just want to know how my xp90C can read 40C at the base with a screw in probe and yet my cpu reads 27C at the same time using the bios or MBM. I also want to know how a cpu and a 1.5 pound block of copper can go from 35-45C almost instantly to room temperature and don't tell me your 120mm fan does the job. I strongly suggest you use some other form of measurement to verify your temps. Or go on as you have been and believe what ever your bios or MBM tell you.


  15. Do not, I repeat do not trust software and motherboard sensors to give you real temps. They are so far off it hurts. This isn't just with DFI but with most motherboards. Didn't you ever wonder how with a DFI board that the temps jumps 10C instantly and drops instantly when the load is applied and removed. The reality is that the temps climb pretty slow and usually reach the first peak about two minutes into prime and when the load is removed the temp drops very slow and usually takes up to five minutes to get back down near the point they started at. I use a laser thermometer to check all of my temps and I have a set of probes for my fluke 87 for hard point temp monitoring. I really want to know what the chip is doing and not just a software estimate.

     

    And no there is almost no way your chip will idle at 25C think about it, applied voltage at idle will create more heat then most room temps.


  16. My FX-60 will do 3.0 but requires 1.5 volts to do so. It will run 2.9 at 1.425 volts and 2.86 at default voltage. Using air cooling all I will say is this thing is a heat monster. When you start shooting the juice to it you get some serious heat out of it. I have settled into the 2.86 13x220 and am very happy with how it runs 24/7 for me. Now for benches and other playing it will boot at 3.15 without a problem and run just about anything I have thrown at it but refuses to run prime at that speed, but the most voltage I have put to it is 1.55 volts and it gets rather warm in a hurry.

     

    It might also just be luck but the memory controller in this FX-60 is so much better than the three dual core opterons that I tried before this. None of the 175's that I tried liked high htt and refused it with 4 sticks of 512 while the FX-60 has run 290x10 with 4 sticks. ( http://valid.x86-secret.com/show_oc.php?id=72982 ) Many have posted good luck with the dual core opterons reaching these speeds but I tried three and the best I could get with default voltage was 2.5. So I'm not sure which is cheaper in the end to just buy what you want or to gamble and not get what you were after.

     

    http://www.v-rodforums.com/member/bench/3150.JPG


  17. Keep in mind that no OC is a sure thing no matter which chip or stepping you purchase. It is still largely luck to get a really good chip. Another thing to keep in mind is the 165 is limited to the 9x multiplier meaning it takes a healthy HTT running all the time to get 2.7 even if you get one of the really good chips.

     

    Good luck no matter which one you decide on.


  18. Since my case is so small, I only have room for 3 fans. My HSF is a 110CFM, my exhaust is also 110CFM, and my Front intack fan is a 150CFM Delta.

     

    Even in a small area changing the fans directions with deflectors or other items can be very important to getting the air moving in the areas that you want it to move. Don't be afraid to spend some time with different methods of moving air around. I play a lot with small deflectors as changing the air stream changes everything. Even a vortex generator at the exit of a fan can get the air spinning and help the flow get where you need it.


  19. Fan location and direction can be different on every case. I live in a very warm area and watch temps like a Hawk. I have several XP90C that work well but what I have found is they tend to hold the heat around the motherboard in the tower cases that I use. For my FX-60 I went with a Thermaltake tower 112 that uses six heat pipes but the fans never blow air near the motherboard. Instead they are aligned so that the cooler exits right into the discharge fan for the case. What this does is exhaust the hot air out of the case instead of blowing it down on the motherboard. Still quiet but I run at idle at 42C and full load of prime at 52C peak after hours with the room temp at 70-75F. Getting the hot air out of the case is very very important and what ever you do don't trust the temp readings in the bios or I should say you cannot trust them on any of the three DFI boards that I own. Did you ever notice they peak temp and drop temps almost instantly, well you can bet your cpu isn't peaking and dropping that fast and my laser thermometer confirms the fact that my cup is climbing slowly and dropping temps slowly. I found one of my boards to be off by as much as 10C when at full load, but the fact that they show the instant increase and decrease got me checking my temps using other methods.

     

    Bottom line is to play with your fan locations and check your temps with something besides the bios or software with this board. Think of it like a car or a bike in the wind and fan location and aerodynamics are everything in keeping the cpu cool.


  20. I picked up a used FX-60 on ebay for just about the cost of a 4800+, I wanted to run 2.8 on a dual core the same as I did with my 146, my 148 and my fx-57 but so far my luck with dual cores had not been as good. I went through three 175's that two peaked around 2.5 and one was terrible and peaked on default voltage at 2.3. Unlike many people the peak to me is 1.4 volts on the cpu, I will not run one higher than that for 24-7 use. I did water cooling and pelters years ago and refuse to put up with that mess for a couple of 100mhz now. The point is the FX-60 that I got will run 2.85 on default voltage, sure it will run 3150 with some boost and will prime at 3.0 with 1.5 volts but I refuse to run one that way so I stick with 2.85 and am very happy with the performance of this chip. I know there are several people out there that get opterons that run that fast however after three failed attempts to reach that speed it was in the end cheaper to just buy the chip I wanted to run at the speed I wanted.


  21. Those are some good numbers, however are you sure your timings are changing as tight as they show they are? The reason I ask is I have had problems with tweaker really changing the timings in the past and running much much looser timings with TCCD your less than one second faster then I produce on air.

     

    3150tmb.JPG

×
×
  • Create New...