Still an Intel part, if I remember correctly.
Yes, the MTH involved was an Intel part. The problem however was not with the part itself, but that some 3rd party motherboard manufacturers were putting the MTH on too long of a path. Had they followed Intel's instructions it wouldn't have been a problem. The part failed not through a flaw in the part itself, but in the failure to adhere to the specifications that the part required. Still, Intel recalled it anyway, because they had a reputation to uphold.
Regardless, I remember the recall, and you would agree it was kind of a fiasco. (and let's face it, reference mobo's have very little to offer to enthusiasts).
I don't argue with either point there. Though there were non-reference mobos that ran the MTH without flaw, because they at least followed the MTH's specs.
The thing is, nearly everyone blames Intel for this failure, and almost no one realizes that it was actually just a handlful of 3rd party mobo manufacturers that weren't following the specifications that were actually the problem.
And if I recall correctly, even if it was indeed post RDRAM then it was already when RDRAM was dead (which Intel tried to shove down our throats at prices that just did not make sense), so no I can not let them slide on that one.
Actually, it wasn't when RDRAM was dead at all. It was at the very beginning of the use of RDRAM. That was
the first motherboard to use RDRAM. So Intel giving a replacement mobo with RDRAM (that performed significantly better than the cheaper SDRAM version that had the problems) was a huge benefit. It was a free major upgrade. Well, free in cost anyway. Obviously frustration and downtime was the price to pay. True, RDRAM was more expensive, but Intel had nothing to do with that. And it
did perform better than SDRAM. But this
was also nearly six years ago now.
And last but not least, their arrogance in the past couple years with a failing [desktop] design is appaling
Failing? That's highly debatable. You'd have to be talking about the Prescott core specifically, as the Northwood (especially the NorthwoodC) was quite an excellent core, and gave AMD a lot to worry about.
Yes, Prescott was hot. And yes, the performance of Prescott on a clock-for-clock basis was slightly lower than the NorthwoodC. Still,
failing is a rather harsh judgement, as it
did perform almost as well as a NorthwoodC of the same clock, and it
did clock higher than a NorthwoodC. It performed adequately. Not great, not bad. Just adequately. And, in fact, the later models of the Prescott
have improved on the heat output, even though no one wants to seriously report it.
There are several things that I personally don't like about Prescott's design, most specifically the cache latency and the resimplification of the misprediction handling to always use the worst-case scenario. I'd hardly call it a failing design though. Compare it's performance to that of VIA's CPUs for example. Now
there is a failure in nearly every way.
Nor do I see it as arrogance to continue selling it, because obviously it's still making them quite a bit of money, and obviously it's still an adequate (just not impressive) design, so it's good business sense. We as enthusiasts might not want to buy one, but then we
do have other choices. For most people out there, it does meet their needs. So how is that arrogance?
Nor do I consider it appalling for the same reasons. Also because, in its own way, it has done a lot to promote AMD's excellent CPUs. Would AMD have the market share that they have today had Intel just shrunk the NorthwoodC core, fixed a couple minor things, added SSE3, and doubled the cache? No. It's only through Prescott's being just adequate that allowed the excellence of the hammers to shine.
Sometimes I even almost think that Intel intentionally did this just so that AMD could gain market share and take Intel safely out of the monopolistic limelight. There was a time when people were about ready to slam on Intel for being a monopoly with the same zeal that they were attacking M$. But now a day you hardly hear anyone but the most rabid of fanboys spouting that level of fire and brimstone. Whether intentional or not, it has certainly given Intel a lot of breathing room again, and quite possibly averted a serious crisis. I wonder how much money it's saved them, and how much money they've lost from market share by comparison.
I am not about to call you an Intel fan-boy, but I sense you are being protective of the giant.
No, I'm being fair. There's a difference. When people have their facts wrong, and pass unfair judgement, I provide the correct facts. I'd do it equally for any company. It's only because most enthusiasts have a grudge against Intel and put AMD on some pedestal that I often end up defending wrongful accusations against Intel. Were the tables turned I would not hesitate to do the same for AMD. And, in fact, there have been many the occasion in which I have lauded AMD.
I would turn that around though. Why have you not given Intel a balanced and fair examination? Why do you hold such a grudge against them?
But we usually get back to the original topic in the course of events.
Facebook
Twitter
Instagram