alansupra94
New Member
Which is faster and why? I have both and I am wondering what will be faster?
Thanks
Alan
Thanks
Alan
Intel Core Duo T2400 vs. Intel Core Duo 2 T2390
- Thread starter alansupra94
- Start date
jdbennet
New Member
The core is an older chip. its not 64 bit, so its not futureproofed. Also, its slow as it doesnt have its own memory unit on the chip itself, which also makes it terribly slow for floating point calculations,
The core2 is newer, supports 64 bit, and fixes many of the shortcomings of the original core.
Go for the core2.
The core2 is newer, supports 64 bit, and fixes many of the shortcomings of the original core.
Go for the core2.
TrainTrackHack
VIP Member
Yes, go for core 2, but... what do you mean it doesn't have its own memory unit? All CPUs have some kind of cache, and only AMD CPUs have onboard memory controllers. Also, memory speed is a minor factor as far as floating-point calculations are concerned.
jdbennet
New Member
TrainTrackHack
VIP Member
jdbennet
New Member
Go at it at the point of view of a low level ASM programmer...
The core is slower than the core2 at doing floating point calculations due to the fact that it has only a single FPU and SSE unit. This makes a big difference in games. Due to the fact this single unit is shared between the two cores, and intek chips dont have an onboard MMU, it is very slow because its a bottleneck as everything has to up via the northbridge and then back into the other core. If both threads try to do FP calculations, it grinds to a halt.
The core2 adds more FPUs and SSE units , which reduces this bottleneck. Also it has an extra ALU which speeds up tasks which make use of two threads (many games use one thread to perform one calculation, and another to do a seperate one, then they want to do something with the results. As the core2 has 3 ALUs as opposed to two, this makes it much faster as it allows complex calculations without having to put the result in main memory or on the stack.
Thats why the core2 is substantially faster at calculations.
The core is slower than the core2 at doing floating point calculations due to the fact that it has only a single FPU and SSE unit. This makes a big difference in games. Due to the fact this single unit is shared between the two cores, and intek chips dont have an onboard MMU, it is very slow because its a bottleneck as everything has to up via the northbridge and then back into the other core. If both threads try to do FP calculations, it grinds to a halt.
The core2 adds more FPUs and SSE units , which reduces this bottleneck. Also it has an extra ALU which speeds up tasks which make use of two threads (many games use one thread to perform one calculation, and another to do a seperate one, then they want to do something with the results. As the core2 has 3 ALUs as opposed to two, this makes it much faster as it allows complex calculations without having to put the result in main memory or on the stack.
Thats why the core2 is substantially faster at calculations.
TrainTrackHack
VIP Member
Ah yea, I saw that 
I though you actually meant to say that Core CPUs would suck at FPU calculations BECAUSE of the missing MMU, not because of having fewer FPU units... my bad, my bad.
I though you actually meant to say that Core CPUs would suck at FPU calculations BECAUSE of the missing MMU, not because of having fewer FPU units... my bad, my bad.
2048Megabytes
Active Member
I do believe you are misinformed. Dual-Core processors have four Arithmetic Logic Units, not three.
alansupra94
New Member
Just putting this out there but the T2400 has a faster clock speed than the T2390....I am not sure if that makes a difference or anything....
2048Megabytes
Active Member
From my knowledge the Core Duo T2400 processor is faster than the Pentium Dual-Core Mobile Processor T2390. The reason it is faster is the Core Duo T2400 has more cache than the T2390.
Cache (also called Static-RAM or S-RAM) is a small amount of high speed memory that is much faster than primary storage random access memory. It speeds overall computer performance by temporarily holding data the processor may use in the near future.
Clock speed of a processor (currently expressed in gigahertz or GHz) is becoming less important as processor technologies are being improved.
Something else interesting to know is that gigahertz means 1 billion cycles per second.
Cache (also called Static-RAM or S-RAM) is a small amount of high speed memory that is much faster than primary storage random access memory. It speeds overall computer performance by temporarily holding data the processor may use in the near future.
Clock speed of a processor (currently expressed in gigahertz or GHz) is becoming less important as processor technologies are being improved.
Something else interesting to know is that gigahertz means 1 billion cycles per second.