Pentium IV Hits 2 Ghz

A number of people wrote in with the news that Intel released the 2 Ghz chip. The Tech-Report article points out a couple interesting meta-ideas - this is Intel's chance to retake the performance crown from AMD, as well as being one of those round numbers that makes people feel warm and fuzzy. I'm sure there's going to be gobs of benchmarks today - post 'em in the comments as you find 'em.

Comments

The challenge of large numbers

[gelfling]

Most usability scientists agree that no one can distinguish much of a difference in PC performance 25% greater than the base value. When PC ran @ 200Mhz it was no big deal to squeeze ~50Mhzout of it since that was simply a quality control variable in the manufacturing cycle. Now with 1.4-1.9Ghz PCs you need to squeeze another ~350-500Mhz out of it before anyone notices so difference between old and improved performance. Just to keep pace with perceived performance you have to add nearly 500Mhz - that is, for lower values there is NO perceived benefit. Which translates into people willing to pay roughly ZERO for anything less than a 500Mhz improvemen. ZERO dollars for which
Intel may have invested billions of dollars to generate. You see it's kind of like boiling water. Nobody cares if it is difficult to raise the water temperature to 211 degrees - it's the 720x more energy required to raise the water that last degree. So it better be worth it to you to spend the energy doing it because investing only 600x more energy will not boil the water.

Thunderbird now, Palomino or Northwood later.

[Max Romantschuk]

It's cool that Intel hit the 2GHz mark, but all that clock speed is really going to waste for the moment.

Right now, you should go for a Thunderbird (AMD Athlon). Later on a Palomino (AMD next-generation Athlon) or the upcoming Northwood (0.13 micron Intel P4) is a better option.

Am I just saying this? No, take a look at this.

It doesn't matter

[wiredog]

I notice very little difference between my new GHZ machine and the 333 MHz machine it replaced. Compiles run faster, but I spend very little time compiling. I spend most of my time editing, and the processors have been able to keep up with my typing speed since the days of the 486-25. Web surfing? I/O bound. Video output? Also I/O bound. Most everything is I/O bound these days. Bus speed is more important than processor speed today. After all, when was the last time you saw anyone discussing spreadsheet recalculation performance?

enough to heat a small home

[juventasone]

detailed review with benchmarks at extremetech.com

I'm curious where power supply requirements are headed. A year or two ago, 230-250W was fine, now I'm seeing Intel and AMD demanding 400W. The HFCs that come with these things are now two or three times the size of the socket. With PCs outnumbering vehicles (saw that stat somewhere) I wonder how the power demands and the heat generated will effect global warming and such.

Sure, its probably not much more than a few light bulbs right now (in both aspects). But like I said, where is it headed.

Re:But really, what's the difference?

[Mike1024]

Hey,

is there really a big difference between 1.9 and 2.0 on the software that people use today?

Well, that would depend on what you are doing. If you were, for example, word processing, you would notice practically no difference, since for the majority of the time, the processor is not being fully utilised anyway. In word processing, bottleknecks are more likely to occour from a program being slow to load (i.e. hard disk speed), or the fact Microsoft Word sometimes likes to move things on the page around for what seems like no reason at all.

If, however, you are doing a highly processor intensive task, like rendering a 3D scene in Caligri TrueSpace 4, you would (in theory, at least) notice a reduced render speed, if you cared to time it, because the processor is being used extensively in the rendering operation.

The problem with this, as with many things, is that the ultra-high-end chips are almost always disproportionately expensive. A 2Ghz chip will likely cost more than twice what a 1Ghz chip costs. Furthermore, a second-hand processor takes a big price hit, so staying 'bleeding edge' isn't really an option. If you have enough money to upgrade every time a new chip comes out, you have enough to get a rendering cluster, which will be faster.

So, where will a 2Ghz chip find a market? Firstly, among 'Power-stupid' people. They will buy ir because hey, it's... like... TWO gigahertz, which is twice as fast as a one gigahertz chip. They likely won't actually need the power, but they have more money than they know what to do with, and iw will be good to brag about.

Secondly, when it's cheaper. As the price drops off, if it can beat AMD's best offerings, people looking for high-end systems will like it.

Thirdly, corperate types who were considering making the switch to AMD because the performance was so much better. If Intel can beat AMD's performance, then AMD will be less attaractive because the performance isn't better, and 'Nobody ever got fired for buying Intel'.

Just my $0.02

Michael