Need help on AMD VS INTEL for programming

K

ktk_ace

Limp Gawd
Joined
Apr 10, 2005
Messages
446
My friend recenty got money to build a com and he was asking if an AMD K8 3000+ 939 pinned cpu based computer is better or worse at writing AND compiling/Running the following programming languages than an intel P4 3.2G lga 775 :

c

c++

Java

Visual Basic

Assembly

and another thing: My friend is picky about the level 2 cache as in P4 1mb vs K8 512 KB.
Does it affect compiling programming languages???

a third thing: he is quering about ddr2 533(Pentium4 3.2 lga775 prescott) which cost Twice as much and its performance benefits over ddr 400 (K8 Athlon 64)...

comments and real life experiences compiling Java and C++ with Athlon64 and P4 machines will be welcomed as on which one is faster and by how much.
 
If I remember correctly, Intel chips have a history of compiling the Linux kernel faster than AMD chips. This might have changed with the new K8 chips, I don't know. In reality, it's not a major difference that it would steer me into buying one over the other.

In terms of running programs, that depends on the type of program you're running. It's possible that the AMD chip performs faster, but it's entirely possible that the Intel chip performs faster. It just depends on what you're doing.

As for the programming languages, they are platform independent, except for assembly. Intel and AMD do share an instruction set, however there are differences between the two. For example, the AMD64 chips do support some 64bit instructions. These, of course, are not available on 32-bit Pentium 4 chips. So, again, it just depends what you're friend wants to do.

Good luck,

Mr_Superstar
 
Its all the same as far as programming goes as it should be. Tell your friend to buy whichever system his wallet is happy with. As for ddr2 I dont see it being a worthwhile investment... a ddr2 chip at same frequency / timings as a ddr1 chip performs the same.

/me votes amd
 
Honestly the small difference in compile time will not be worth worrying about. The difference in compile time between my work machine with a 2Ghz pIII and 512mb of ram and my home machine with a 3.3Ghz PIV and 1gb of ram is nothing worth noting and I work with a wide variety of projects. I have projects that take seconds to compile and one that takes roughly an hour to do a complete rebuild.

As far as what affects compile time. Everything affects compile time honestly. For large C++ projects the biggest bottle neck I notice is hard drive performance. I can't speak to the other languages since I don't use them enough to say.

Intel machines do usually make better compile machines in my experience because of their better stability. But I'm an Intel fan so I'm biassed.
 
In my experience, the higher clockspeeds of the netburst arch (P4) doesn't improve compile time significantly. Compiler code is branchy and integer intensive and doesn't benefit from MMX, SSE, SSE2, 3DNOW, etc...

In practical terms, I have noticed that a P4 1.6Ghz and a 1Ghz Athlon with comparable memory, disk, etc... will compile with very similar speed. Disk speed will probably affect compile times more, and some build tools will compile several files simultaneously, which can reduce whole-project compile time significantly.

For real (if dated) benchmark data, see here.

Linux kernel compile times may also interest you, but I didn't find good comparisons for my 30 seconds of effort.
 
There really is no difference.

I have done lots of work on AMD rigs, and on my current Intel setup, and have found no discernable difference between the two.

Both will work just fine.
 
You must log in or register to reply here.
Back
Top