Jump to content

Bman

Members
  • Posts

    6
  • Joined

  • Last visited

Bman's Achievements

Newbie

Newbie (1/14)

  1. Someone better tell them they need to track down those loose cannons that are publishing this secret information on the AMD web site, then. Odd. In the article at that sample link I gave, they claim that the Toledo core has about 230 million transistors in a 199 mm^2 die, whereas the Manchester has about 154 million transistors in a 147mm^2 die. I wonder if they got their facts wrong... I had to sacrifice somewhere... (a couple somewheres, actually.) However, to be honest, the 4200+ has plenty of horsepower for my needs. Also, overclocking was not really on my radar at the time I made my decision. Some people are claiming that the Manchester core overclocks better than the Toledo, and some are claiming the opposite. See this thread: http://forums.overclockers.co.uk/showthread.php?p=5826260 Interesting... On that page (and the associated details page) they list the voltage for the 4200+ as 1.35-1.40 V. So, I guess I can consider myself to be running at "stock" voltage at 1.40 V. I wonder why when the CPU voltage setting is set to "Auto" in the BIOS it uses 1.30 V... Thanks again for all your great input!
  2. It may be the pack rat in me, but I'd prefer mine to live longer than that. Also, 3 year old CPUs are far more useful than they used to be.
  3. I do realize that everything is relative. While 0.01 volts is not much to us (although, current comes into the equation too...) we operate at a scale somewhat larger than nanometres. I'm fairly certain that the X2s also use the Manchester core (that's what CPU-Z reports and a Google search seems to confirm this). Here's a sample link: http://techreport.com/reviews/2005q3/athlo...00/index.x?pg=1 It appears that at least some versions of the X2s that have 512k L2 cache use the Manchester core, which is physically different from the Toledo. I did read a posting that said there are also X2s with 512k L2 cache that are Toledos with half the cache disabled. Thanks very much for the info. As an update, I seem to be Prime95 stable at 2420 MHz at 1.4V. The 10% overclock was symbolic to me for some reason. I don't seem to have a great overclocking CPU here, but I was damned if I wasn't going to get 10% out at a conservative voltage. There are people on the forums claiming significantly higher overclocks on stock voltage for this CPU (or the 4400+).
  4. I test using two instances of Prime95, with affinity on one set to CPU0 and affinity on the other set to CPU1. The instance of Prime95 running on CPU1 is always the one that fails. I'm running the torture test, with the amount of memory each instance can use restricted to 400 MB (I've got 1 GB total memory) so that both cores will run at 100% (I believe Angry_Games pointed out how to do this). Just by way of some more info, as my cooling is currently lacklustre, for my testing I've got the side of my case open with a 14" fan blowing air onto the motherboard. This keeps my reported CPU and Chipset temps below 40C and my PWMIC below 45C. Right now, I'm just trying to determine what would be achievable if I had better cooling. Although it's unique, I don't think I want the open case and 14" fan to become part of my sig. Thanks for the feedback.
  5. Mine is an E4 Manchester. I believe the stock voltage is 1.3 V. At 1.3 V, I can only get a 5% overclock. I did get it prime95 stable at 2585 MHz (235 x 11) at 1.6 V. However, after reading about the long-term effects of higher voltages, I don't want to keep the voltages that high. I'm currently testing at 1.4 V, and that seems as though it might give me about a 10% overclock. In all cases, I'm held back by my CPU 1, which seems far less happy overclocking than CPU 0 (which seems to be a common experience with X2s).
  6. I'm new to DFI and to overclocking. I've been reading through the forums and gained a great deal of insight -- thanks to all who post and to DFI for supporting it. There seems to be general agreement in the postings that increasing the voltage to your CPU can shorten its life. However, beyond that, there doesn't seem to be much consensus at all. Some seem to feel that any voltage above stock will shorten the CPU's life -- it's just a matter of how much. Others seem to feel that as long as you're under 1.5v (for a 90nm core), you're not going to noticeably impact its life span. Others are comfortable jacking up their voltage into the stratosphere as long as their reported core temps stay within a decent range. (Several people have argued, though, that voltage alone can kill a CPU regardless of temperature.) Does anyone have any authoritative input on this issue, or know where such input can be found? This leads me to another question: does increasing the voltage on your memory, chipset, and LDT bus shorten their life spans also?
×
×
  • Create New...