Jump to content


  • Content Count

  • Joined

  • Last visited

Everything posted by IncredibleHat

  1. Vid: 2.6 ? 2.7 ? For memory? What are those numbers from? I am confoozled. Do you mean 2.7 volts? 2.7 GHZ (thats impossible with ram hehe)? 2.7 CAS? 2.7 mean 270? jcarno: What does your SPD report those ram as? If 3-5-5-10... I would be greatly interested in ALL settings you have for your BIOS on those!
  2. I found mine to be stable at 250 with the 3-4-4-8 setting. 25 passes of memtest5, and just an hour test in prime95 (I was getting tired of tests). Mine are in orange. I wanted to be able to run at 300FSB (x9 2700mhz cpu) with a 9/10 divider (270mhz ram). However it looks like that is not anywhere near possible with this particular ram. I am going to have to drop it all the way to a 5/06 divider (246mhz ram). Maybe I can get my cpu stable at a few more FSB... need a new HS fan first. Pfft. I'm beyond the RMA period for newegg. eBay time! QUOTE: I would just be somewhat desillusioned if I can't overclock this RAM. I hear you there... from a company who touts themselves as having great ram for OCers... they really sold a pile of horse droppings with these!
  3. I would like to add as well, I am in the group who got the "bad" SPD reporting type ram as well: SPD: 250 3 5 5 10 I have a feeling they started to sell some crap there the past month or so... it seems those who have gotten good results from the HZ, bought theirs long before this 'bad batch' came about? Would be interesting to see on the XLS chart "when" people bought their G.Skill.
  4. Same. DFI Official 623-3 bios. I have more luck though... I can get it to post. Even when I have set each and every item in the list. Actually, I have very few things on auto. However, my problem is getting it stable anywhere above stock 250 clock. There are some evil settings in there that wont let me get beyond POST, but most settings allow me to get to memtest. My report image I have shows a bunch of settings I tried that let me get to memtest (but not stable)... as well as a few at the end that simply wouldnt let me boot. SETTINGS GRAPH
  5. I will add my results here too. Well, not for the xls file (which is awesome btw!), but more of the "yeah, these new HZs are performing horribly for me too!" I wrote up a big report over here: EOC Thread - My Report Basically, I ran into a wall at 260. I am actually going to see if these are even stable at 250. I know they are stable around 200, as thats what they were at during my CPU OCing attempts! But 200 ?!?!? GAH! Really aweful ram as far as I have. Other people may have better luck with theirs, but I am running out of ideas!!! :confused:
  6. The stock heatsink on the Opteron 165 is a serious device that rivals some of the best replacements. Basically, you don't need to replace the heatsink... but the fan. I am still using the stock fan, but I put one of those ducts in the middle of it to eliminate the 'dead spot' a fan makes on a heatsink. I have been hunting around for a decent 80mm fan to replace the stock one with... that isn't terribly noisey like the tornado. Twivel: I think you mean F instead of C? Hehehe ... OMG 80c in your room!!!... live in arizona? Hehe. My room is ambient 24.5c most the time during the winter.
  7. To help make that more clear, here is a list: Total Voltage - Idle Temp. - Load Temp. v1.30 - 32c - 37c v1.35 - 32c - 37c v1.40 - 32c - 39c v1.45 - 35c - 42c v1.50 - 36c - 44c v1.55 - 36c - 45c v1.58 - 37c - 49c v1.65 - 38c - 53c The total volts is mostly a mixture of VID and VID Special. Most of them above 1.4 I used a 110% VID Special Control. When I set my VID to 1.35 or 1.4, I had the Special set to "Auto", which had an effect of making the actual vcore lower than the setting. So like my 1.35v * Auto, was 1.32v. Kind weird... so I stopped using auto.
  8. I may have the guts, if I can find a way of cooling it down from that evil 52c... but I am already above where I had simply wanted to be, which was 290. Lapped is a process of sanding down the rough stock surface of your heatsink to make it a mirror finish. By doing this and using "Arctic Silver 5" as the thermal paste, you can reduce your overall temps by up to 7c in some cases. I dropped mine down only 4c... so I may re-seat the heatsink again, or try lapping it more. My mirror finish wasNT prefectly mirror, had a wee bit of fog to it, cause I didn't spend the extra hour on the 2200 grit sandpaper Right now its been holding on full dual primes at 45c. Which is very nice for me.
  9. Welp, can't get it stable at 312 ! Heh. Prime95 flat out throws errors seconds into core2. My core2 has been giving me issues through the whole process (well, not big issues, its just the first one that has a problem so I never get to a point where core1 has a problem, heh). So, I am now prime95ing on 300x9 @ 1.425v*110% (which is 1.54v in MBM5/BIOS). Appears to be holding steady on Large TFT. I tried to get 310 going on prime95... no luck. Unless I boosted the vcore to 1.50*110% ... which made MBM5 scream bloody murder at me, as I have my alarm at 52c, and it hit 53c, so I stopped prime95. It was holding though! Just WAY too hot! Super Pi did not get it anywhere near that hot, but then super pi isn't as stressful. So, instead of dinking with the max heat/max fsb/max vcore, I dropped it down to the above 300 settings, and will see if its prime stable by morning. At least that will give me a good "OK THIS IS GOOD" point.
  10. Apparently not. As soon as I had to raise the vcore, it shot up like a rocket with each step of 5. I am at 312, because 315 is proving impossible, no matter what vcore I give it (temps remaining under 50c as well). So, I backed down one by one and found a super pi stable spot at 312. I am noW testing 312 with other apps.
  11. I got tired last night at 280-285... it failed super pi at 285, but a more humane failure, so it is noW time for me to raise vcore above 1.30v. So I thought I would go to bed and start again today =)
  12. Yeah, I got the tip on another site about my BIG FOOKEN MISTAKE with the Max Async Latency, and Preamble settings. WOOPS! I am now well past that 265 wall, and working my way up the scale. *phew* !!! I about dropped a brick in my pants ! Thanks for all the feedback though! I wrote down some of those ideas and numbers. I have dropped my LDT to x3 now, since I am beyond that 265 zone around the 2000 (1000x2) mark. Good deal... thank goodness it was something silly like a small setting I screwed up on, and not something serious! EDIT: Forgot to add my stepping: CCBWE 0551 VPMW
  13. Well, I started a thread on another forum about this, but thought since I am a DFI user, I may find more help here. So here goes: I began the process of ramping up my rig (see sig). First step always is "Find the max of your CPU". So, I drop my ram to 1/2 divider, set LDT to 4x, and start cranking away at the FSB. I get to 260 and it passes the rudimentary super_pi 32m tests. I go to 265, and it doesn't get to even the "Verifying DMA pool data..." message on pre-boot. Or it DOES, but windows promptly dies when it reaches the desktop and MBM5 tries to boot up (BSOD). So. 260 sounds awefully pathetic. After much adjusting various settings, I found that although 260 is stable-ish... if I change the ram divider to ANYTHING but 1/2... it will fail to even get to windows (stuck before that DMA message). This leads me to believe I may be having an issue with my ram. Its good stuff... came highly recommended by opteron owners and dfi users. And passed initial memtest testing overnight on stock settings at stock cpu speeds. What I have set in my BIOS for ram: CPC: on TCL: 3 TRCD: 4 TRAS: 8 TRP: 4 TRC: 14 TRFC: 16 TRRD: 2 TWR: 3 TWTR: 2 TRTW: 3 TREF: 3120 (setting this to 3684 results in a scrambled pre boot screen) TWCL: 1 Interleave: on Skew: auto Skew: 0 DS: 5 DDS: 1 Async: 6 Response: normal Preamble: 5 Cycle: 128 Dynacount: off Bypass: 16x Max: 5x Gran: Off/4 Voltage settings: VID: 1.30 VID+: x104% LDT: 1.2 Chip: 1.5 Ram: 2.6 The above is all the same when I moved from FSB 200 to FSB 260. Using a LDT/FSB of x4, and a divider or 1/02. Like I said, when at 260, if I move the divider t any other setting, it no longer will reach boot sequence. Help? :confused: I have read the ENTIRE guide... even have it printed out. I even read the big guide over on short-media. I dug through many posts here before I started OCing, and I just cannot figure out what in blue blazes is going on. It feels like my ram is going wonky. Or if feels like the whole IDE bus is falling apart at high speeds, or something else is going on. ?
  14. Most people recommend just one dab of arctic silver in the center, and letting the heatsink press it outwards when you clamp it down. It's what I do. A teeny dab, about the size of a BB. Arctic Silver 5 isn't dirt cheap, but one little hypo will last you eons.
  15. Which is always good! Silence is golden... AMC can't be wrong, right?
  16. Dude, you got some serious cleaning to do. I would never let my system get to that point! AGH! AGH !!! MY EYES!!! I feel filthy now
  17. Theoretically, no. However it will reset the RAID/SATA settings in the BIOS to "disabled", so you will need to go "enable" those drives under BIOS BEFORE you attempt to boot off the raid.
  18. The book shows this for the PWR-LED, and made me wonder "why"? My old board (an MSI) only had a 2-pin. Am I jumping to conclusions when I think that this 3-pin means I could hook up a dual color LED to it? That it would somehow show Stand-by color and ON color or something?
  19. Coke in a can for me! Anyone else peeved that they "Put Vanilla Coke on Hiatus to sell their god AWEFUL Black Cherry Vanilla Coke?!" Vanilla coke in the can was the drink I looked forward too each day to start work with. Now I have to mix my own... pfft!
  20. QUOTE: "Jack I suggest you use something besides software to check your temps." Well, that wasn't very helpful. Jack, first you should figure out what your ambient temperature is in your room around the case. Make sure you figure it out for celcious, makes it easier. Any thermometer will work, just let it sit for a good hour or two before reading it. When you figure out your ambient, add about 4c to that number. This should be your "inside case ambient" temperature to aim for (or below). Then, from there, you would want to add about 2c to the inside case ambient temperature, and that should be your CPU's IDLE temp. Basically, your CPU should not idle more than 6-7c above ambient room temperature. If you are, then your case needs modding to get it cooled down inside better. For figuring out the LOAD temp of your CPU... that varies on a lot of factors. A "good" range is no more than 10c over idle. A decent range is 15c. Anything more, and you may need a new HSF combo, or lapping, or some arctic silver 5 thermal compound. All of those things will help keep your CPU at a lower temp, especially under load. Big rule of thumb though: Keep your CPU below 50c at all times. Especially if you are overclocking. If it just aint possible... then you may have to move to more radical cooling methods, OR lower the overclock settings.
  21. Although the hotfix is a start, its also worth noting that playing any game for long long periods of time, usually results in a bomb of some sort. That's just windows for you However, if every game you play does this, and within an hour or two, then yes, it could be something else in the mix besides windows being an butt. Many times in the past, on a machine I KNEW was hardware-stable, I would get bombs in various games after playing them in long spurts (like 3-4 hours). I had just adopted a method of "save, quit, reboot, load, and load saved game". I recall a few games that were notorious for crashing out after extreme periods of play... most liekly memory leak type issues. Something you have no control over. Just keep that in mind. It could very well be a bad card, or it could just be bad programming in the game and/or windows. Try the hotfix, and let us know!
  22. 31c load? That is impressive, what's your ambient? (I assume thats priming on both cores)
  23. Wow... so we just have to guess at what the things in BIOS do. Hrmpf. Ok. Consider the idea dropped. :drool:
  24. Closest thing I keep coming too is: http://www.phoenix.com/en/Customer+Service...nts/default.htm But that page sucks for useful information. Maybe I never found that rojakpot site because I didn't use the keywords of "optimize". Didn't think to use it, heh... cause I figured a outline of definitions is just that, a list of definitions. Gah... google wasn't my friend today!
  • Create New...