Personal computing discussed

Moderators: renee, Flying Fox, morphine

 
GodsMadClown
Gerbil First Class
Topic Author
Posts: 123
Joined: Thu Feb 27, 2003 10:02 am

Damage & crew? Doing a review of the 90nm 939 chips?

Wed Sep 29, 2004 1:41 pm

If so, these are some good questions that I'd love to see answered.

http://www.hardforum.com/showpost.php?p ... stcount=34

Ok, clock for clock, does the 90nm CPU actually outperform the 130nm CPU? What if the 130nm CPU had the 1 meg cache?
 
SpotTheCat
Gerbilus Supremus
Posts: 12292
Joined: Wed Jan 29, 2003 12:47 am
Location: Minnesota

Wed Sep 29, 2004 2:37 pm

have you heard of NDA?
 
sativa
Grand Gerbil Poohbah
Posts: 3044
Joined: Sun Apr 14, 2002 7:22 pm
Location: lafayette, la

Wed Sep 29, 2004 2:40 pm

SpotTheCat wrote:
have you heard of NDA?


they are already on sale i think. and i've seen reviews already.
Science is forbidden. Laboratories manufacture danger!
 
Damage
Gerbil Jedi
Posts: 1787
Joined: Wed Dec 26, 2001 7:00 pm
Location: Lee's Summit, Missouri, USA
Contact:

Wed Sep 29, 2004 2:59 pm

Yeah, they're out there. We'll see.. ;)
Scott Wasson - "Damage"
 
anand
Gerbil Elite
Posts: 901
Joined: Tue Mar 25, 2003 12:30 pm
Location: Dallas, TX, USA
Contact:

Wed Sep 29, 2004 3:10 pm

Are there even any architectural differences between the 90nm and 130nm parts? There was a big difference in the P4 because the Prescott had some major changes in the design along with the process shrink.
 
Damage
Gerbil Jedi
Posts: 1787
Joined: Wed Dec 26, 2001 7:00 pm
Location: Lee's Summit, Missouri, USA
Contact:

Wed Sep 29, 2004 3:34 pm

Nope. This is just a die shrink.
Scott Wasson - "Damage"
 
Usacomp2k3
Gerbil God
Posts: 23043
Joined: Thu Apr 01, 2004 4:53 pm
Location: Orlando, FL
Contact:

Wed Sep 29, 2004 3:51 pm

bummer :wink:
it's new enough, that they shouldn't really have many changes
 
GodsMadClown
Gerbil First Class
Topic Author
Posts: 123
Joined: Thu Feb 27, 2003 10:02 am

Wed Sep 29, 2004 7:25 pm

http://www.anandtech.com/news/shownews. ... ART=21&p=2

Thanks to Buck_Naked from our forums to alerting me to this information on AMD's 90 nanometer Athlon 64 processors (there a link to benchmarks as well):

* Full SSE3 implementation
* Improved hardware data prefetch mechanism
* Increased number of writing combine buffers (D0 stepping A64's can now combine up to four non-cacheable streams compared to 2 o­n the C0 and CG stepping A64's)
* Improved o­n-die memory controller with more advanced open page policy
* O­n-die thermal throttling
* Black Diamond Low-K technology (slower less power hungry transistors in less used sections and faster and more power hungry transistors in frequently used sections of the cpu)
 
sativa
Grand Gerbil Poohbah
Posts: 3044
Joined: Sun Apr 14, 2002 7:22 pm
Location: lafayette, la

Thu Sep 30, 2004 2:07 am

GodsMadClown wrote:
http://www.anandtech.com/news/shownews.aspx?i=22530&ATVAR_START=21&p=2

Thanks to Buck_Naked from our forums to alerting me to this information on AMD's 90 nanometer Athlon 64 processors (there a link to benchmarks as well):

* Full SSE3 implementation
* Improved hardware data prefetch mechanism
* Increased number of writing combine buffers (D0 stepping A64's can now combine up to four non-cacheable streams compared to 2 o­n the C0 and CG stepping A64's)
* Improved o­n-die memory controller with more advanced open page policy
* O­n-die thermal throttling
* Black Diamond Low-K technology (slower less power hungry transistors in less used sections and faster and more power hungry transistors in frequently used sections of the cpu)


has this stuff been verified? sse3 support?
Science is forbidden. Laboratories manufacture danger!
 
Forge
Lord High Gerbil
Posts: 8253
Joined: Wed Dec 26, 2001 7:00 pm
Location: Gone

Thu Sep 30, 2004 4:01 am

I should know very shortly and will be happy to confirm/deny.
 
rika13
Gerbil Elite
Posts: 720
Joined: Tue Sep 30, 2003 9:01 am
Location: Normal, IL
Contact:

Thu Sep 30, 2004 5:27 am

i have heard they get hotter, seems to do with the smaller die having less area to dissipate heat is what they are saying
ASUS P7P55D-E, Core i7 860, EVGA GTX 570 SC, 4GiB DDR3 (OCZ I think), Razer Lycosa (letters rubbing off) and Lachesis, Benq G2400W, Steelseries Siberia G1 headphones (no speakers)
 
GodsMadClown
Gerbil First Class
Topic Author
Posts: 123
Joined: Thu Feb 27, 2003 10:02 am

Thu Sep 30, 2004 7:22 am

It has also been rumored that the new 90nm 3000+ and 3200+ do not have these features, and that they are reserved for the 90nm 3500+. Just telling you what I heard. I'd like to know if the 3000+ support sse3. Are there ways to test for the other features mentioned?
 
sativa
Grand Gerbil Poohbah
Posts: 3044
Joined: Sun Apr 14, 2002 7:22 pm
Location: lafayette, la

Thu Sep 30, 2004 12:14 pm

GodsMadClown wrote:
It has also been rumored that the new 90nm 3000+ and 3200+ do not have these features, and that they are reserved for the 90nm 3500+.


I doubt this. AMD bins its chips & having 2 different chip designs wouldn't be very cost effective..

they wouldn't just 'throw away' chips that have sse3 support, but dont mee the 3500+ rating. Then again, they could disable sse3 support somehow and label it a 3200 or 3000 :-?
Science is forbidden. Laboratories manufacture danger!
 
UberGerbil
Grand Admiral Gerbil
Posts: 10368
Joined: Thu Jun 19, 2003 3:11 pm

Thu Sep 30, 2004 12:50 pm

sativa wrote:
has this stuff been verified? sse3 support?
There's certainly nothing I've foud on the AMD site to verify this. Their most recent Tech Doc (Sept 14th) is a builder's guide that specifically mentions SSE2 but not SSE3.

Which is not to say it's not possible -- they may be emargo-ing their tech docs until all the NDAs are lifted.
 
Damage
Gerbil Jedi
Posts: 1787
Joined: Wed Dec 26, 2001 7:00 pm
Location: Lee's Summit, Missouri, USA
Contact:

Thu Sep 30, 2004 12:56 pm

Is TR the source of this report? :)

http://techreport.com/onearticle.x/6363

We are trying to confirm whether these changes are present in current 90nm Athlon 64s.
Scott Wasson - "Damage"

Who is online

Users browsing this forum: No registered users and 1 guest
GZIP: On