Is a desktop Pentium 4 2.8 ghz faster than a Duo 2 Core 2.2 ghz laptop chip?

I am just wondering because I have a Pentium 4 chip that is about 3 years old in my desktop rated at 2.8ghz, and I just bought a new T61 with the Duo 2 Core 2.2 ghz chip in it, and don't understand how a new chip is only 2.2 ghz and my 3 year old chip could be faster at 2.8ghz?

 

Speeds is less important than it used to be the architechture of the chip has changed. Your Core Duo2 is faster (it also has two cores on one die) making it faster.

 

Not to mention about 2-3 times faster than your P4. The P4 archeitecture is inferior the the C2D's.

 

The Core microarchetecture is faster then Netburst(that your P4 has) also, C2Ds have 2 cores making them able to process 2 things simultaneously.
Note: C2D are quieter & more efficient also.

 

Faster clock speed does not equal faster CPU. Look at the AMD vs. Intel chips, for example...AMD's chips generally use much lower clock speeds than Intel, but still maintain similar overall performance. Architecture optimizations aside, keep in mind the Core 2 Duo has two CPU cores on a single die (P4 only has one), resulting in significant performance improvements for any multiprocessor capable applications. It also has more L1 and L2 cache, and is 64-bit, which doesn't really speed it up, necessarily, but makes it capable of running 64-bit operating systems. You also need to factor in that front-side bus and memory clock speeds have also been increased since the P4.

Try to find some benchmarks comparing the two with each other, I think you'll find that your 2.2 Ghz Core 2 Duo handily outperforms your 2.8 Ghz P4.

Clock speeds aren't as useful as they were 5 years ago in comparing CPU's.

 

Unfortunately. I remember when it used to be so easy to tell what CPU was better. Now it's much more complicated.

What everyone else said, P4s can't compare to C2Ds, even mobile C2Ds.

 

Sometimes I am proud of the knowledge of NBR members. Nothing to add.

 

You, these discussions lose a lot of distinctions

There's a lot of confusion between cpu clock speed, arcihtecture, cache size and how long it takes to execute a job mix?

Job mix? Yes job mix. Is a give piece of code single threaded or multi-threaded.

If it's single thread then the 2.8 gigahertz pentium (hyper-threaded or not hyper-threaded?) may be faster. Are there other processes running> A duo-core can handle those in the other core. What's the cache size, not mention the relative memory speeds? How about disk rotation speed and form factor?

The question is so complex that the question itself almost doesn't make since.

 

When Pentium M's were released a few years ago the general rule of thumb was PM's were about 50% faster than equivalently clocked P4's. Core CPUs are even faster than PM's (multiply that times 2 for two cores).

So a 2.2GHz C2D x 1.5 (or higher) x 2 = P4 speed equivalent.

 

With modern CPUs, the limiting factor in your computer is the hard disk. I can't wait until SSD goes mainstream.

 

even that's not enough. The buses that transfer this info are still slow. Need to get the memory speeds equal to the CPU speeds, graphic speeds etc. Be nice to see a computer with no bottlenecks.

 

"With modern CPUs, the limiting factor in your computer is the hard disk."

This has ALWAYS been the case.

But this discussion still lacks a lot of precision. The question really cannot be answered without more information because the answer is going to depend more on the load than it will the computer system.

 

I think you need to go back to Renee's comment about what is being processed. In general Dual Core is going to be faster, but it isn't about pure speed. Its about the parts of the system integrated to solve the problem of executing instruction sets of increasingly complexity faster, or for that matter multiple threads at once. The question which is faster is assuming we are doing an apples to apples comparison, and instead its more like comparing the turbo prop versus a jetliner.

 

The busses now are 1000+ GHZ although they aren't that fast in a thinkpad.

Here's secret... technology is just going to move the bottlenecks around. There will always be bottlenecks.

As a developer, I'm very happy with the speed of a thinkpad as a second system. It's more than able as a development machine.

 

Yes, Phil. You have to specify the kind of load, before one says what system will process that load the fastest. Is it memory bound? Is it I/O intensive? If so, the processor won't be the critical question, it will the disk. Is the test localized in memory or is thrashing all over a large memory space? That's another situation where the disk will determine which finishes sooner.
It is multithreaded or single threaded. Is the pentium in question hyperthreaded? If the load is single threaded and highly localized then it will all be processed in cache so the relative cache sizes will be the determinant.

 

I used to have a 3.2 gigahertz hyperthreaded pentium in my desktop.

I now have a dual core extreme (2.93 Ghz). There are many, many situations where that 3.2 Gigahertz hyperthreaded pentium will finish faster than my dual core extreme.

 

I have a quad core, and a lot of software isn't optimized for 2 cores let alone 4...so again I concur some software is slower on it -- but in general it is quite a bit more robust than my P4 EE.

 

I am a performance nut although I am not a gamer. But I wouldn't buy a quad core at this time for exactly the reason you stated.