JC, Re: "> I think the Pentium 4 is far better equipped than
> the Celeron in handling new kinds of tasks.
That's a very subjective statement. If we ignore the memory bandwidth advantage, then I am not sure that I agree with you. MMX is finally in compilers and is getting into binaries. And P4 suddenly halves the max throughput of MMX compared to an equally clocked Celeron."
No, in some case, latency is slower, but throughput is not. In fact, throughput is increased due to higher clock frequencies.
Re: "P4 has smaller data cache than a Celeron."
And how much of a penalty do you figure this causes? Do you have any data on the miss rate before and after the changes?
Re: "It has L2 cache that on tests appear to have higher latency than the L2 on the Celeron."
Absolutely not true! The L2 cache is lower latency on the Pentium 4, and the bandwidth is much larger as well. Since cache bandwidth is directly related to frequency, the Pentium 4 has an even larger advantage. The Pentium 4 cache also has the ability to do more more loads and stores per clock as the P6.
Re: "Celeron can decode more instructions per clock than Pentium 4."
And this doesn't help much if these instructions aren't getting executed. You also forget that the majority of instructions in the Pentium 4 get accessed from the Trace Cache, NOT the decoder.
"Does this sound like a processor design that is uncontestably better made for the future?"
Yes! You apparently do not fully understand the micro-architecture, which is why you are making these lame attempts to show that Intel's 5 year old P6 micro-architecture is somehow more advanced than their brand new Netburst micro-architecture. If you really knew all the changes that went into the Pentium 4, you would see that it is far more advanced, and is much more capable of higher performance than any P6 core, especially the Celeron.
"Or is the only real benefit to the memory heavy assumptions of times to come just the higher frequency and bigger memory bus?"
No, the Pentium 4 has many features that increase IPC. However, more than any micro-architecture in the past, there are also a lot of micro-architectural performance issues. Many research firms, as well as Intel themselves, have discussed some of these issues. They are reported in technical journals, at trade shows, or off the cuff, but I am aware of quite a few things that are in the progress of being vastly improved. Because you have concentrated more on benchmarks than understanding the actual micro-architecture, you have fully underestimated it. I don't blame you, but in the future, you will be very surprised with what the Netburst core can deliver.
wanna_bmw |
