|
|
#1 |
|
Hypermodder
Join Date: Nov 2009
Location: Romania
Posts: 678
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Thuban(X6) vs Deneb(X4) single-threaded performance?
Can someone who is very familiar with these two series confirm or correct my assumptions below?
This is about single-threaded performance only!
Where does 'Turbo Core' fit into all of this? Can it tip the scales into Thuban's favor? |
|
|
|
|
|
#2 |
|
I Endure
Join Date: Jul 2004
Location: Jasper, AL
Posts: 4,426
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
The Thuban cores are a bit more refined than Deneb, and therefore artificially created single thread performance favors Thuban. It's part of the enhancements along with Turbo Core, which automatically overclocks a core or cores in response to increased load.
|
|
|
|
|
|
#3 |
|
Hypermodder
Join Date: Nov 2009
Location: Romania
Posts: 678
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
OK, but what if you were to disable Turbo Core (it can be disabled, right)? Would the performance be identical for chips running at the same frequency?
|
|
|
|
|
|
#4 |
|
The modding head is firmly back on.
Join Date: Mar 2010
Location: Gliding through cyber space
Posts: 7,322
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Thuban will be slightly better with or without turbo core . Once you start overclocking turbo core becomes pretty point less anyway as more often the overclock is running higher than the turbo although it can be set to run at a higher frequency.
|
|
|
|
|
|
#5 |
|
Supermodder
Join Date: Jan 2011
Location: China :(
Posts: 317
![]() |
Thuban is like deneb 2.0 + 2 cores......
in other words Thuban is a improved version of deneb with 2 more cores.... the improved architecture therefore renders a slight increase in single threaded performance..
__________________
Main rig: CPU: Phenom II 955 @ 4ghz cooled with a H50GPU: HD 5850 Powercolor (Bricked) awaiting new card to hold me out till AugustMotherboard: Gigabyte 890xA-ud3 RAM: 4gb ddr3 1600 OCZ @1600 |
|
|
|
|
|
#6 | |
|
Multimodder
Join Date: Dec 2011
Location: English Riviera
Posts: 171
![]() ![]() |
Quote:
It's not a repeat of Northwood vs Prescott, thankfully, the new Thuban architechture is more efficient than Deneb. Pity we can't say the same for Zambezi, eh? |
|
|
|
|
|
|
#7 |
|
Supermodder
Join Date: Jan 2011
Location: China :(
Posts: 317
![]() |
although this is true... looking at overall performance its an interesting attempt for a whole new concept on a CPU.... if the Architecture develops i reckon it could have potential...but for now. it can go get bulldozed.
__________________
Main rig: CPU: Phenom II 955 @ 4ghz cooled with a H50GPU: HD 5850 Powercolor (Bricked) awaiting new card to hold me out till AugustMotherboard: Gigabyte 890xA-ud3 RAM: 4gb ddr3 1600 OCZ @1600 |
|
|
|
|
|
#8 | ||
|
Multimodder
Join Date: Dec 2011
Location: English Riviera
Posts: 171
![]() ![]() |
Quote:
An interesting concept, Hyper Threading but in Hardware. But then the Sinclair C5 was an interesting concept, too, but it wasn't any good, was it? Quote:
|
||
|
|
|
|
|
#9 |
|
Hypermodder
Join Date: Nov 2009
Location: Romania
Posts: 678
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
OK, thank you for your input.
I would also like to know if there are any benchmarking tools for single threaded performance, other than Super_PI. |
|
|
|
![]() |
| Thread Tools | |
|
|