Silicon or Succotash, Memory Conflicts among three 875/865 motherboards.
Succotash is a traditional staple-food among the Pennsylvanian Dutch. Consisting of boiled green corn and lima beans, sautéed onion, bacon, and served in a cream sauce. It's what one makes, when the crop's fail, and one is desperate. What does this have to do with the memory conflicts I'm about to detail? About as much as the memory itself. I'm beginning to believe those early, nefarious rumours, describing the Springdale (865) chipset as being conceived in a conspiracy between Intel's marketing, and engineering. Claims have been made, Q&C tests performed on initial batches of 875 silicon (NB-MCH) where the NB silicon failed so many criteria, and a budget version had to be released. Intel, in an effort to save thousands of pieces from the dumpster, announced Canterwood's younger brother, Springdale. Has Intel taken sub-par silicon, and made Succotash? I sincerely hope not.
I do pray the problems detailed below may in fact be BIOS related, and therefore fixable. No matter to Intel, because those most affected are Overclockers. A company which locks its multiplier internally is not necessarily looking to cater to the Enthusiast/Overclocker market, which comprises less then 10% of sales. Intel didn't grow into the manufacturing giant they are, by meeting the fickle needs, of a few hundred-thousand Overclockers. The chipset does, for the most part, perform very well. And even with its memory divider idiosyncrasies, it's currently the best out there. Regardless, it's becoming crazy-making to re-set the BIOS by literally having to "pull the plug", remove affected (or infected) memory, and re-set the BIOS by replacing newer DDR with older DDR memory (I've been using a 256MB stick of OCZ GolD3700, I'll explain why this works). It's been quite a long time since I recall being so frustrated with PC-hardware. This article details the anomalies I've experienced among three motherboards, and several of the latest memory's designed for the Canterwood/Springdale chipsets.
I'll also discuss some theories as to the origin, and one company's memory, which initially criticized, has left a few critics with shoe laces dangling from the sides of their mouths. When using certain memory, beyond a certain speed, and/or with any "dividers" on the Abit IC7-G, Abit IS7-E, and Epox 4PCA3+ the boards become "locked." One becomes unable to re-set the BIOS via the jumper less mode, shutting down and powering up or even clearing the CMOS! Unfortunately both Corsair (Twin-X4000 (512MB-kit/1GB-kit)), and OCZ Dual Channel GolD4000 (1GB-kit)) are most adversely affected. It's becoming next to impossible to benchmark these products due to the anomalies. An experienced Overclocker performing a thorough test must lower the FSB in 1MHz increments when seeking to establish, the module's highest stable speed. Now imagine having to cut the power, pull the memory, replace with a 256MB stick of older DDR, clear CMOS, plug-in, re-start, re-set all BIOS values, enter Windows, and finally shut down through the OS (Operating System). Then repeat these steps each time certain FSB-speeds, memory dividers, or memory timings are chosen causing another lock-out? Slightly inconvenient.
The most egregious of the anomalies occurs on the Epox 4PCA3+. Whenever the above conditions cause a lock-out, one is totally unable to re-set BIOS values via conventional methods. This is not only beyond inconvenient, but the constant powering down of the PC by literally having to "pull the plug" is bound to have detrimental long-term effects. And not only on the micro-circuitry, DIMM slots/clamps, etc., but the CPU as well. When all is said and done, I can't remember a chipset as fickle as the 875. It's painfully tedious to find a given hardware's true performance limits. This adversely affects my ability to write, as it threatens deadlines, and indefinitely prolongs the review process. Why I'm the first to write about this, is beyond me. I will not be providing many benchmarks in this article, because it is about failure, not success. If, however; you would like to read about my results with all these products and many more, please feel free to visit my website: Brainstroms
and it's Review Pages
You'll find a plethora of info pertaining to Intel and AMD overclocks, as well as other hardware experiences.
Just so you know, as a student of Philosophy, the validation of empirical data (repeating the experiment) is the cornerstone of my methodology. I understand how variables (or not accounting for them) can render any conclusions false, even if the premises "seem" to be true. Based on the number of times I've had to remove memory, and replace it with another to re-set the BIOS, I'm either the Galileo Galilee or Aristotle of motherboard BIOS testing. Take your pick. I only hope like Socrates, Intel doesn't force me to ingest poison for criticizing their chipset.