"Hyperthreading" on Athlon64 X2?

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Jeff7181

Lifer
Aug 21, 2002
18,368
11
81
It most definately does not use hyperthreading. The bit they're referring to in the OP's link is in reference to the CPU's multithreading capabilities. Hyperthreading is a form of multithreading... as is dual cores. As you know, an HT enabled P4 shows up in Windows as two logical processors... dual core processors will show up in Windows as two physical processors. In both instances, the CPU is basically just telling Windows, "I'm capable of executing multiple threads."
 

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
Originally posted by: mamisano
Did anyone actually read the article? All it means is that the Dual Athlon will enable the Hyperthreading bit on the chip. This way, when a program written with Hyperthreading support is run, it can take advantage of the 2nd core. It does NOT mean that AMD has included hyperthreading, they just want to take advantage of programs already designed for multithreading via Hyperthreading.


Exactly and this has been talked about for a few days now in other forums...
 

carlosd

Senior member
Aug 3, 2004
782
0
0
Now with dual cores HT is useless. Besides AMD doesn't need HT, they think it is a waste of resources, HT is needed by P4 CPUs to keep up with A64 CPus in the tests, because of defficiencies in P4 architecture. A64 CPUs are very efficient and will take no advantage from HT.
 

Jeff7181

Lifer
Aug 21, 2002
18,368
11
81
Originally posted by: carlosd
Now with dual cores HT is useless. Besides AMD doesn't need HT, they think it is a waste of resources, HT is needed by P4 CPUs to keep up with A64 CPus in the tests, because of defficiencies in P4 architecture. A64 CPUs are very efficient and will take no advantage from HT.

HT might have improved the inefficiency of the Windows scheduler when multi-tasking, pointed out in the article on the main page. It's not just the P4 architecture that HT benefits... it just happens to be the most widely known. I believe IBM also uses a form of HT, but not because of a rediculously long pipeline, but because of a rediculously wide pipeline. When execution units go unused, HT can potentially improve performance by making use of them. I think the reason AMD decided not to use a form of HT is because maybe the K8 architecture is well rounded enough that the performance gains wouldn't be worth the R&D to put it into production... especially when they have dual core plans.