Outline ·
[ Standard ] ·
Linear+
:: AMD Phenom ::, Barcelona Launched, Phenom Next..
|
kmarc
|
 Jul 28 2007, 09:46 PM
|
The future is here - Cryptocurrencies!
|
QUOTE(stevenlee @ Jul 28 2007, 09:37 PM) last time with 2000 HT it beat pentium D and now 4000HT, AMD might think can over take C2D as C2D is almost 2x faster than PD It doesn't work that way. Even 1st generation HTT of 1000Mhz is already too much as reviews showed that a HTT of 200-400Mhz is more that enough for everyday gaming/usage. IIRC, even Intels new FSB 1333Mhz is in preparation for quad core computation. I don't think the new HT3 of 4000Mhz will lead to any real world benefit as a HTT of 1000Mhz is not even fully utilized! |
|
|
|
|
|
kmarc
|
Jul 28 2007, 10:33 PM
|
The future is here - Cryptocurrencies!
|
QUOTE(kmarc @ Jul 28 2007, 09:46 PM) It doesn't work that way. Even 1st generation HTT of 1000Mhz is already too much as reviews showed that a HTT of 200-400Mhz is more that enough for everyday gaming/usage.
QUOTE(stevenlee @ Jul 28 2007, 09:59 PM) that true for now... but we dont know the main purpose they increase the HT.... cos computation basically did in procesor and seldom at memory....thus 4000HT wont benefit at all..but how if they implement the 4000HT as a link between the core and the data might transfer faster between 2 core and compute data ... that just my guessing...the real 4000HT still haven tested.... Pls note the phrase "for everyday gaming/usage". AMD's main purpose for HT3 is for SMP/multi-core processing. Here's a link for more info (the part on HT3): http://www.anandtech.com/showdoc.aspx?i=2768&p=3This post has been edited by kmarc: Jul 28 2007, 10:34 PM |
|
|
|
|
|
kmarc
|
Jul 28 2007, 11:19 PM
|
The future is here - Cryptocurrencies!
|
QUOTE(intune @ Jul 28 2007, 10:58 PM) Do take account why the HT3 is developed at the first place..true its to take advantage of the multiprocessing power of the new Phenom architecture but also to give hardware vendors and manufacturers the access to create expansion card and also accelerators card that could tie in directly into the cpu itself. For e.g the AGEIA PhysX card which utilises bidrectional communication from Gpu and CPu before frame is rendered. Unlike game console the bottleneck for PC's are between the CPU and GPU. The Phenom is clearly engineered to take advantage of the SMP/multiprocessing server market other additional features are also tailored of future gaming GPU architecture cards too.. All will be revealed in coming weeks about perfomance and numbers as Phenom launches.. As for me im only interested in price.  Are you talking about HTX? Is it the same as HT3?  I do sincerely hope that this technology can be used for expansion/accelerators/physx card (in terms of gaming/daily usage). Anyway, here's some other links to clarify on HT3/HTX connectivity: http://news.soft32.com/hypertransport-30-r...ublic_1324.htmlhttp://www.hypertransport.org/tech/tech_htthree.cfm?m=3http://www.hypertransport.org/tech/tech_htx_main.cfm?m=9Note the quote for the first link :"The new HTT 3.0 will be excelent for computers with more then 2 processors, because for 2 processors the available HTT 2.0 it's enough. HT3 makes sense for socket F because the HT link IS a bottle neck in 4-way+ servers." This post has been edited by kmarc: Jul 28 2007, 11:20 PM |
|
|
|
|
Quote
0.0168sec
0.60
6 queries
GZIP Disabled