Dear Scumbria:
There are a number of ways to take the report of a non working P3 1133 MHz.
1) Its a bad CPU. FALSE! Tom claimed that it ran in all platforms at 850 MHz. It just means that he received a badly binned part. This would not be good for Intel Q/C and/or P/R. (Paul: Are you yelling at Intel for shipping a badly binned part?)
2) It needs a microcode update to run 1133 MHz. Possible. It did run at 850 MHz but, not at the claimed speed. If it takes a motherboard microcode update, many upgraders and people with a favorite board will not be able to use it. (Paul: Why did Intel not post what platforms it could run on? AMD has an approved motherboard list.)
3) No microcode update is needed. Someone at Intel paniced, and lied about what was needed. Bad Q/C and Bad P/R, BIG TIME!
4) Bad CPU Heatsink thermal contact (overheats then shuts off). Again, bad Q/C. Lower clock does work.
5) Intel supplied motherboard failed to show up. Company used to ship overseas (Tom's is in Germany) lost motherboard & memory. They should have shipped early to allow for eventual foul ups and allowed time to take corrective action (Made sure he got working stuff for review on time (Again bad P/R and/or Q/C).
6) Anand and Sharkey got top speed CPUs while Tom's is more typical of production (passed all tests but was marginal). This could be worrisome for the future. 2 good out of three shipped is very bad error rate (Intel's reputation will be flushed if, this continues).
7) Tom's CPU was dropped/shocked/etc during shipping (bad packing. I once got memory from Mushkin in an inflated static free "penumatic" shock resistant package (esentially a spherical air bag with the DIMM suspended in a pocket within)). Maybe previous person tied to overclock or other such bad activity and pemanently down binned the CPU.
8) CPU has a lifetime problem. It runs at 1133 MHz for a period of say two weeks and then slows down as time progresses. The higher voltage may cause a shortened life and this CPU was good but is now not fast enough to make it after retests. This is the assumption with all overclocked parts. You hope that the overclocked part lasts until you update it with the next "greater" part. This may be the worst of all cases for us the consumer. If it is ever found out that a CPU run within spec will routinely do this before the warranty is exceeded easily (that a 3 year part does not last for 3 years (even worse less than a year)), that manufacturer will get hit with a lot of "Badwill" and many law suits (and deservedly be run out of business). I hope this is not the case.
9) Tom is stupid. Not possible. He has many reviews (some even Paul points to) that make him a bonifide reviewer. I think that many blame the messenger when their "baby" is shown in a very bad light. I give him the benefit of the doubt (he has much to lose and not anything to gain by these type of review). So, unless you can prove he is wrong, these results must be accepted as facts.
Many other even more wild speculations are possible with this limited sample (only three sites reviewing it???).
IMHO, 1 or 2 are the most likely. Intel is to blame if, either are true.
Tom tried five different boards, one was MIA (Bad P/R in itself), one supposedly had the "fix" (Not something we like to hear), and the last three were test beds used in other reviews that have been positively pointed to before by those whining now. All failed at 1133 MHz and four ran at 850 MHz (MIA can't run at any speed but zero). (Paul: You jump on us if, the roles were reversed. So take your medicine and yell at the ultimately responsible party, Intel)
Pete |
