Missing The Boat
I’ve been looking at these Northwood vs. Athlon reviews. Unfortunately, lots of data, not much useful information.
There is no point generating reams of data for just configurations nobody will use. No real hardware enthusiast in his right mind is going to
buy a $600-$700 2.2GHz PIV processor and then just run it at spec. Either he’s going to overclock it or decide to buy the XP instead.
So what he wants to know is 1) how far can this be overclocked and 2) how well does it do against the XP when it’s overclocked? That’s what he wants to know, even if he’ll never overclock. He needs to get an idea of the potential of this new CPU.
Having the overclocked AND spec results can actually help a bit, but just having spec results doesn’t help our enthusiast much at all.
He reads these reviews, though, and he might as well be reading PC Magazine for all the answers he’ll get to his questions.
It’s a little more likely people will buy an XP2000+, but even here, an extra $100 is hardly worth the extra you get, and again, cheaper options with just a little overclocking should get you to the same level.
We’re Not All Made of Money
What is important to the average, real person is not the extra 3% or 5% or 7% in performance being laboriously recorded, but what the damn things cost for those small improvements.
True, there will be a few for whom cost is no object, but I’d say that’s at most 5% of the readership, if even that.
You would think the average reviewer might note that Intel and AMD are trying to get back to the bad old days of sky-high pricing, and that this is bigger news to the average real person than a few percentage points of increased performance.
But no, we generally get silence. Even worse, we get “let them eat cake“ statements like, “The Pentium 2.2GHz will cost a bit more . . . .”
Might I suggest that the current $350 difference between the two is not precisely a mere trifle to the average reader? May I notice that even the price of the “cheap” XP2000+ is quite a jump over even recent CPU pricing? Dare I say that maybe, just maybe, the small amount of extra performance you’ll get from these isn’t worth the large amount of extra money required?
Could I state that maybe more than a few places have lost touch with their audience, and that when Intel and AMD are trying to get back to the bad old days of sky-high pricing, this is not the time to go silent on this subject?
Or maybe it’s just us.
Two Types of Overclocking
There’s always been two types of overclocking: “hobbyist” overclocking and “economic” overclocking. Hobbyist overclockers are this generation’s hot-rodders; they just want to push a piece of equipment to the absolute max.
“Economic” overclockers (and you certainly can be both) have historically been more motivated to get the most for the least amount of money.
In the recent past, the need for “economic” overclocking has dropped quite a bit simply because the price differences between (primarily AMD) processors has been small enough to not be worth the bother to many.
Well, the case for economic overclocking is coming back, but you’d never guess it from the reviews.
Treating Overclocking Like A Dirty Word
This is the feeling I get from these reviews. Some places don’t even bother anymore.
One place tells you that overclocking the FSB is really a bad thing to do (this is extraordinarily bizarre considering the website practically invented it). After saying that was the reason why they didn’t even try to overclock a PIV system, they then proceded to do just that with their AMD system. To quote the article completely out-of-context, “this makes little sense.”
Most places stuck a reference to an overclocking attempt somewhere in a back room of their article. For instance, this place dedicated a whole one sentence in a thirteen page article to it. So far, I’ve only run across one place that was nice enough to give you even a little idea how well the overclocked CPU actually performs at the overclocked speed. Perhaps not so coincidentally, that was the only place that indicated their PIV came off-the-shelf.
Unless I’m completely out of it, seems to me that would be of far more interest to the typical hardware enthusiast who might actually buy one sometime soon than forty-three reams of data saying what it does at spec.
I suppose we should be a little thankful for the crumbs showing how far the various parties were able to get, though the details that would indicate how hard they tried are usually scarce or not present.
Are We Now That Far Out On the Fringe?
Maybe I’ve just lost it. Maybe this site is out on the lunatic fringe nowadays.
Are we crazy for noticing and not liking these prices? Are we mad for openly suggesting these products aren’t quite worth the price tag? Are we lunatics for wondering why overclocking has become the crazy aunt in the attic?
Or are we just sociopaths for talking about things that everyone else has decided not to talk about?
We got into doing this looking for high-performance, value computing. We saw overclocking as a tool to reach that end. When pricing policies changed, we considered that more a victory for value computing rather than a defeat for overclocking.
We thought our beliefs represented those of at least a large portion of if not most computer enthusiasts.
It seems, though, that if we look at the other sites, the world has changed and passed us by. Has it? Have we lost touch with our audience? Does all that matters nowadays is getting on the gravy train and getting that review with a wall of numbers out the minute the NDA expires, and sounding like rebels but acting like good little boys?
Our problem is not that we’re doubting ourselves, but rather that we’re not. We still think our beliefs are the beliefs of much if not most of the audience. That means somebody’s out of touch, and we’re just asking for a sanity check to make sure it’s not us.
So which is it?
What We See Out of All This
We find both the XP2000+ and Northwoods overpriced for what they deliver. Unless all you do is play Quake, we don’t see any reason why you would want an unoverclocked Northwood over an XP.
Based on very limited data, at least the initial 2.2GHz Northwoods are very unlikely to run at 133Mhz FSB, or 2.933GHz without extreme measures being taken. It seems more likely the 2.0A GHz Northwood might be able to reach or get close to 133Mhz FSB, or 2.67GHz.
Even there, again based on very limited data, it looks like reaching 2.67GHz is going to be tough using reasonably conventional means, and it won’t be an easy, or maybe even very safe overclock. We’re doing our homework before buying this, and at the least, you should crib from us before you do so, too.
Although we’re going to do it (once we can find a 2.0A at a somewhat reasonable price), we don’t find Northwood overclocking particularly compelling at this point. The CPU costs too much, and even if it gets to 2.67GHz; it’s still not going to blow the XP away, especially from a price/performance standpoint.
We don’t expect to see no-brainer Northwood overclocking (by that we mean an easy 33% overclock with little effort) until the next stepping. By that point, we’ll probably see the first .13 micron Thoroughbred available. How those two will compare overclocked is anybody’s guess at this point.
P.S. You’re Just Jealous, Why Don’t You Do Real Work, etc., etc., etc. . . .
Some of you may say, “This is just a poor substitute for some real work.”
Well, in all honesty, it’s not like anybody is going to be in a real rush to buy this stuff given what it costs. I’m sure we could get on the lists to get preproduction samples, but given what these reviews all look like,
we have to wonder if ours would be any better if we had to follow the same set of rules.
There is no lack of places able to generate all the benchmarks one’s heart would desire. What seems to be in scarcer quantity are places that ask the sort of questions and raise the kind of points we raise. So we do tend to think more and test less than most places, but that’s just because we find less of the first than the second, and you need both.
Per doing real work, the real work wasn’t done, that’s the whole point of the article. Something that looked like work was certainly done, but it didn’t answer the real questions the average person visiting these places would want to know.
Maybe real work can’t be done anymore, at least not within the timeframes and restrictions imposed.
Or maybe we’re just nuts.