Did CPU speed increases slow down?

[glen]

?

 

[dmcowen674]

Probably wrong forum area but yes the speed race has been over for two years now after AMD kicked Intel's butt with slower processors.

 

[Savij]

Yes, we've been going multithreading then multicore for a while.

 

[glen]

Originally posted by: Savij
Yes, we've been going multithreading then multicore for a while.

In PC for Dummies language?

 

[swtethan]

Originally posted by: glen

Originally posted by: Savij
Yes, we've been going multithreading then multicore for a while.

In PC for Dummies language?

instead of doing one thing at a time.... several things at once

 

[pray4mojo]

I think I've noticed this too. Ive been out of the loop for a while. Whats the fastest processor on the market?

 

[dmcowen674]

Originally posted by: glen

Originally posted by: Savij
Yes, we've been going multithreading then multicore for a while.

In PC for Dummies language?

The processors are more powerdful than we can possibly use or throw at them.

End of race.

 

[swtethan]

Originally posted by: pray4mojo
I think I've noticed this too. Ive been out of the loop for a while. Whats the fastest processor on the market?

pentium 3 1000mhz

 

[everman]

No I don't believe speed increases have slowed in terms of computation power.
If by "speed" you mean megahertz then yes that particular number is not increasing as quickly because we are getting more out of each clock cycle.

 

[cKGunslinger]

Originally posted by: glen

Originally posted by: Savij
Yes, we've been going multithreading then multicore for a while.

In PC for Dummies language?

I think that was PC for Dummies. You may need it in terms of "PC for Grandmas."

 

[jagec]

Originally posted by: dmcowen674
The processors are more powerdful than we can possibly use or throw at them.

End of race.

So not true. Ever tried rendering a movie?

 

[Fenixgoon]

architecture >>>>>>> clock speed

 

[her209]

Answer me this... why is it that the CPU manufacturers even make chips of differing speeds? In other words, does it take more work to make a chip that performs at 3GHz versus one that performs at 3.4GHz (assume they are the same core)? Its all a ploy to charge us more money for faster CPUs, I tell you!

 

[jjones]

Just out of curiosity, do multi-core systems boot faster?

 

[mercanucaribe]

When I was a senior in high school I had a 1.6ghz cpu. Now I'm a senior in college and I have two 1.6ghz cpus. Go figure.

 

[everman]

Originally posted by: her209
Answer me this... why is it that the CPU manufacturers even make chips of differing speeds? In other words, does it take more work to make a chip that performs at 3GHz versus one that performs at 3.4GHz (assume they are the same core)? Its all a ploy to charge us more money for faster CPUs, I tell you!

The way CPUs are manufactured, in any given batch you produce some will have more flaws than others. The best ones will run at the highest speed, while others are only stable at a lower speed.

Added into the mix are processors capable of higher speeds, but are sold at a lower speed in order to fill a market niche. This is why you can buy a lower speed cpu, and overclock it sometimes to very high speeds.

 

[mugs]

Originally posted by: dmcowen674

Originally posted by: glen

Originally posted by: Savij
Yes, we've been going multithreading then multicore for a while.

In PC for Dummies language?

The processors are more powerdful than we can possibly use or throw at them.

End of race.

Haha... ok. You are clearly wrong.

The speed race has slowed down though, I believe due to heat issues with higher clock speeds. That's why dual-core processors are becoming more popular.

 

[jman19]

Originally posted by: dmcowen674

Originally posted by: glen

Originally posted by: Savij
Yes, we've been going multithreading then multicore for a while.

In PC for Dummies language?

The processors are more powerdful than we can possibly use or throw at them.

End of race.

So very wrong, lol.

 

[Eli]

I'm sure we would all hear about it if Moore's law failed.

 

[Frostwake]

If by "speed" you mean megahertz then yes that particular number is not increasing as quickly because we are getting more out of each clock cycle.

So true... whats really funny is, when conroe comes out later this year and supposedly "wipes" everything else on the market, say i get a 2.4 ghz one... my current 3 year old p4 is also 2.4 ghz... how come in 3 years the ghz remain the same yet the performance increases are drastic? i mean a p4 2.4 vs conroe 2.4 is like night and day... i think this is the first time a company actually walked backwards? current top end p4 go for 3.8 ghz or so i think.. fastest conroe available will be 3.3.. hmmmmmm weird stuff

 

[jagec]

Originally posted by: jman19

Originally posted by: dmcowen674
Probably wrong forum area but yes the speed race has been over for two years now after AMD kicked Intel's butt with slower processors.

So very wrong, lol.

How is that wrong? Two years ago AMD WAS clearly faster, and with slower processors to boot. It was only after Intel started building its own slower-clock speed processors that they began to gain that lost ground.

 

[Eli]

Originally posted by: Frostwake

If by "speed" you mean megahertz then yes that particular number is not increasing as quickly because we are getting more out of each clock cycle.

So true... whats really funny is, when conroe comes out later this year and supposedly "wipes" everything else on the market, say i get a 2.4 ghz one... my current 3 year old p4 is also 2.4 ghz... how come in 3 years the ghz remain the same yet the performance increases are drastic? i mean a p4 2.4 vs conroe 2.4 is like night and day... i think this is the first time a company actually walked backwards? current top end p4 go for 3.8 ghz or so i think.. fastest conroe available will be 3.3.. hmmmmmm weird stuff

Fastest I've handled is the 3.73GHz P4 Extreme Edition.

All the future products are in plain white boxes and say only "Intel Confidential" on them, we don't get to see speed or specs.

 

[thecoolnessrune]

Attention focused from doing one thing as fast as possible (from earliest CPUs to Athlon 64 and Intel A and B classes) to managing more than one thing (Intel HT on C-E classes) to doing more then on thing at once at a slight speed sacrifice (AMD X2 and Intel 8-900 series)

 

[jman19]

Originally posted by: jagec

Originally posted by: jman19

Originally posted by: dmcowen674
Probably wrong forum area but yes the speed race has been over for two years now after AMD kicked Intel's butt with slower processors.

So very wrong, lol.

How is that wrong? Two years ago AMD WAS clearly faster, and with slower processors to boot. It was only after Intel started building its own slower-clock speed processors that they began to gain that lost ground.

Actually I responded to the wrong dmcowen post

You can see my edit above...

 

[Savij]

Originally posted by: glen

Originally posted by: Savij
Yes, we've been going multithreading then multicore for a while.

In PC for Dummies language?

Instead of doing 1 operation at a time very quickly, intel and amd have been moving to doing more operations per cycle to get a net gain in performance.

PS you're on ANANDTECH'S forums. You should really check out the website cause it goes over this stuff really thoroughly. I recommend starting with the 1999 Athlon articles.