These technologies were design to handle Pentium processor’s use multimedia applications more efficiently, each new technology came out every 2-3 years due to updated technologies and new standards for efficient computers

MMX

MMX was design by Intel in 1997 and interestingly enough MMX is not an acronym at all, although these names have come up; MultiMedia eXtension or Multiple Math or Matrix Math eXtension with the last one being coin by AMD during a court battle with Intel at the time. However, two problems would come up in its design which would help produce the SSE technology; the first is that it used integer operations and it reused floating points and so it make it difficult to work with SIMD at the same time.

Now this site provides the best overview of how the MMX technology is used including benchmark tests that been on done on it. It is interesting to note that MMX technology is still being supported by Intel and they mention on their website that processors with MMX technology improves Processor Performance by 10%-20%, and so it would seem that a technology that is 10 years old is still supported to this day.

SOURCES

http://en.wikipedia.org/wiki/MMX
http://www.intel.com/design/intarch/mmx/mmx.htm
http://www.tommesani.com/MMXPrimer.html

SSE

SSE (Streaming SIMD Extensions) would come out two years later in 1999 to do two things, bring direct competition to AMD’s 3DNow and fix the problems of MMX floating point reuse and not relying just on integers to work, instead SSE uses SIMD, but still supports the floating point. During the time of SSE being manufacture it would spawn off 4 more versions SSE2 SSE3, SSS3 (enhancement of the SSE3) and SSE4. Also it is interesting to note that from SSE2 on the P4 used this technology.

SOURCES

http://en.wikipedia.org/wiki/Streaming_SIMD_Extensions
http://www.x86.org/articles/sse_pt1/simd1.htm
http://partneradvantage.sun.com/protected/...t/tech/sse.html
http://en.wikipedia.org/wiki/3DNow%21

Hyper-Threading

Hyper-Threading is basically Intel’s top of the line technology that was introduced to P4 processor and is more advance then the super-threading technologies that were used at the time. In short it improves processor performance under specific work loads, and it is interesting to note and I am only guessing but hyper-threading could be the beginnings of dual core processors since it turns one processor into two processors.
The way Hyper-Threading works is that it “duplicating certain sections of the processor—those that store the architectural state—but not duplicating the main execution resources.” This way the computer can double the tasks it can perform. However, it would seem that Hyper-Threading has a problem and that it is energy insufficient as that it uses 46% more power then the dual core processors.

The following articles give in insight on threading technologies:

http://arstechnica.com/articles/paedia/cpu...erthreading.ars
http://www.intel.com/cd/ids/developer/asmo-na/eng/196575.htm
http://www.hardwareanalysis.com/content/article/1557/

SOURCES
http://en.wikipedia.org/wiki/Hyper-threading
http://en.wikipedia.org/wiki/Super-threading

May I remind you that these days the technology you mention is available in Athlons also. Not sure if all of it, but most of it is. Also a simple explanation of Hyper-Threading is basically having the processor utilize its unused resources and there's nothing really grand about it. About the only place you can see improvements are if the programs are actually coded to run better with H-T.

Aah... sse... I remember when sse2 came out - that thing was a lifesaver. Having three systems simultaneously doing integer calculations (mmx,sse,normal alu) at the same time was probably the best thing about the new pentium and athlon releases at the time. - and all this without 64-bit too.

And St. Michael,I'm glad you did your homework on this one, but it seems that a few facts are underrepresented. Mmx boosts computational speed a LOT - but only if you're using it mainly for integer calculations. Oh, and how come 3Dnow! (trademark) wasn't mentioned in light of its own merit? It was a big step forward at the time, even if it was just an AMD thing. Plus, one could always theoretically say that you'll always get more floating-point performance on an AMD chip because of the extension.