AMD 65 nm next month

Printable View

Show 100 post(s) from this thread on one page

09-09-2006, 08:03 AM
Gam3Ra

AMD 65 nm next month

http://www.theinq.com/default.aspx?article=34247

I've get a Core2Duo this month :slapass: :nono:
09-09-2006, 08:05 AM
Supertim0r

conroe still beat the AMD ;)
09-09-2006, 08:34 AM
Dubz

Quote:

Originally Posted by Supertim0r

conroe still beat the AMD ;)

i knew when i clicked on this post someone was going to say this.
09-09-2006, 08:44 AM
Okda

Quote:

Originally Posted by Supertim0r

conroe still beat the AMD ;)

then why done't u stick to the intel section

reallly guys try to avoid following ur needs for a flamware in every thread
09-09-2006, 09:02 AM
ozzimark

Quote:

Originally Posted by Supertim0r

conroe still beat the AMD ;)

i guess you know how well 65nm is gonna clock then? ;)
09-09-2006, 09:08 AM
arisythila

Seems like there are some Intel guys that are :banana::banana::banana::banana: faced... Some dont even have a "conroe", but they wont hesitate to let you know conroe is all powerful. I was reading some stuff in there yesterday, and some Intel guys ragged on this guy because he had a question. Mind you, I didnt post in the intel section, I JUST READ. Matter a fact. I think when K8L comes out and dominates Intel. We should go over to Intel forums and then tell them how much Intel sucks.. What do you guys think?

~Mike
09-09-2006, 09:12 AM
metro.cl

Quote:

Originally Posted by ozzimark

i guess you know how well 65nm is gonna clock then? ;)

there are already some 65nm benchs out there, same thing as am2 on 90nm regarding performance.

so amd will need 3.4GHz to beat conroe x6800XE (maybe a bit more dependeing on the game/bench) + it needs to have more space to oc because a xe can easily go to 3.6GHz.

still good news that AMD is doing things fast to reduce the performance gap
09-09-2006, 09:18 AM
ozzimark

it will definitely be interesting to see. i've gotta try to save up a bit of money to pick up one of those 65nm guys when they come out. :toast:
09-09-2006, 09:30 AM
Supertim0r

bleh, i'm not a Intel fanboy. Just performance fanboy. I would have a via if it was faster. Some people can't admit Intel is faster now. Back on topic now. Yes, I'll buy a 65nm amd when its available. Even if it's not faster. No fanboyism here
09-09-2006, 09:42 AM
marauder16

I hope in December X2 3800+ 65nm will be available in my country cause I'm definitely buying it, only problem is that I'm afraid that shops wouldn't separate 65 form 90nm, unless if there will be difference in price, what do you guys think?
09-09-2006, 09:45 AM
vitaminc

smoke and mirrors...

If its originally planned for December release, it should already be shipping out from the fabs to the backend assembly plants right now.

Die shrinks will grant around 10-15% perfromance increase, so power to AMD. But it's not going to be anywhere close to Intel though.

I am just hoping AMD's marketing won't do some dumb pricing and keep 65nm 3800+ X2 at the same price as the 90nm parts, and Dell and HP better not suck up all of AMD's capacity. :(
09-09-2006, 09:49 AM
JPeitzman

http://img.oc.com.tw/b60799/20069511501413494030192.jpg <--from [H]

Are there any other ES's out at all, I have heard rumers of ones in Japan, but all I have seen is this ^ and one other site with a few pics of one(could be the same proc.).

Quote:

so amd will need 3.4GHz to beat conroe x6800XE (maybe a bit more dependeing on the game/bench) + it needs to have more space to oc because a xe can easily go to 3.6GHz.

I think that a 65nm AMD could hit 3.4-3.5, maybe better since the 90nm ones are hitting 2.9-3.0. We can't really tell though until they come out how they will clock. Like you said, I don't think they will beat a extreme OCed Conroe, but they might close the gap, maybe even close it a lot faster than some people thought. Although they could also have a lower thermal envelope, be clocked the same as the 90nm chips, and still for some reason have a 3.0ghz wall; that would be a bummer huh. I would love to see these come out though and all of a sudden see some AMD's hitting 4ghz.

I am interested to see the quad core K8's. I know that they won't touch Kentsfield, that is K8L's job (hopefully), but it is nice to know that AMD seems to be on track with 65nm and quad core. Maybe if we see 65nm next month we will see K8L duals in early (like Jan. and Feb.) 07; better still I would love to see AMD to a quiet little sneaky release at like Christmass or something, doubt it will happen, but I can dream.

I just hope that this 65nm AMD does clock good and cause a little unrest in the Intel camps. I am tired of the flaming. I am not saying that Conroe isn't awsome, it is, I have considered getting one. There is also no denying that AMD may not touch Conroe until K8L, 65nm really might be to little to late. There is still no need for the flame and the hate. I have no problem with teaseing and pokeing, but some places (not at XS persay) it has gotten out of control. We are all the same, we just make choices, there really is no need for extreme fanboyism and hate/super flaming.

That said I can't wait to see what Big Green has in store for us in the coming months. I can't deny that I don't like AMD, again though I can't deny that Intel is kind of, well, stomping them right now in performance. I think Green will bounce back, and I think they will bounce back in a big way. Besides in a way Conroe users should welcome 65nm and K8L, because competion is what fuels this industry and I hope that K8L and Kentsfield can push it to the edge once more.
09-09-2006, 10:31 AM
_33

AMD 65nm + S939 ???

If they do release S939 parts, then it will have to be good news. Otherwize it's sad news for my DFI Lanparty and DDR1 memory. When you look at what happened with S754, AMD did release a Venice for that socket. So, I don't see why they wouldn't release the new 65nm chips.
09-09-2006, 10:35 AM
thigobr

Could be better wait for a second revision of 65nm(like Venice over Winchester)...
09-09-2006, 10:49 AM
alexio

I'm guessing with the 65nm process 3ghz or less is the fastest K8 chip we will see. I'm pretty sure that if AMD will have something to compete with Conroe performance wise it will not be untill K8L. The only reason AMD is pushing 65nm is for budget CPU's and later on for K8L. Not for an increase in clockspeed anytime soon.

65nm is nice for people looking for reasonable performance at an amazing price though :)
09-09-2006, 10:52 AM
_33

Reading this, it is inevitable that AMD will just forget about S939. If so, well I know where my money will go for my next system ;) .

http://image2.sina.com.cn/IT/h/2006-...8f3857a384.jpg

The slide here clearly states the new chip will be for AM2 (DDR2). So let's not keep our hopes up for the S939, it's in the trash (read garbage) bin at AMD.
09-09-2006, 11:01 AM
serious1

maybe instead of getting a conroe i'll just get a $50 sempron and a good $200 mobo and wait?
09-09-2006, 11:38 AM
vitaminc

Quote:

Originally Posted by serious1

maybe instead of getting a conroe i'll just get a $50 sempron and a good $200 mobo and wait?

I would just get a Conroe box now. It's pretty much a given that all the 65nm parts will be in extremely limited availability and will be marked up worse than Conroe at launch.

BTW, 939 is officially EoL. Stop having wet dreams about 65nm s939 parts. It aint going to happen.
09-09-2006, 01:23 PM
syne_24

yea let s939 r.i.p; change is a good thing. I cant wait for some 65nm benchmarks to come.
09-09-2006, 02:10 PM
Exedy

i want 65nm on my infinity :cool:

yes..the bad new is that 939 is dying ...:(
09-09-2006, 04:57 PM
Biggles

Going from 0.18 micron to 0.13 micron eventually got AMD an increase of 600 MHz in production processors, on a similar architecture. That's the fastest Palomino, the 1.73 GHz 2100XP through to the fastest Barton, the 2.33 GHz 3300XP. That's over a 1/3rd increase in clockspeed through process changes. It didn't happen overnight obviously, given that the Thoroughbred A came first, then the Thoroughbred B, then finally the Barton. But it was still quite an increase. 0.13 micron went as fast as 2.4 GHz if you count the Athlon 64s and Opterons, but that was a whole new architecture.

Going from 0.13 micron to 0.09 micron got AMD an increase of 600 MHz - from the Athlon 64 3700/3800/4000 at 2.4 GHz to the Opteron 256/856 at 3 GHz. Not as much this time, but still 25%. Again, it didn't happen immediately, there was the revision D processors (basically Winchester) before we got to the revision E chips (Venice, San Diego, Troy etc).

So using previous form, I expect rev F - the 0.065 micron processors - to get to around 3.75 GHz, in production processors. Given that there has always been overclockers who have managed another couple hundred megahertz out of the chips, I see 4 GHz being plausible. It may take another revision to get to that stage, but I do see it happening. Funnily enough, in talking to some process engineers, we reckoned AMD would get about 3.8 GHz from a die shrink, and that was before we did the quick arithmetic.

In light of that, Intel likely will remain top dog since they have headroom with Conroe. But AMD are at least going to be competitive again, and on performance, not just price. Not that they really care if the relatively few enthusiasts have been lost to Intel, they are still selling everything they can make and that is before counting the Dell sales.
09-09-2006, 05:21 PM
metro.cl

Quote:

Originally Posted by Biggles

Going from 0.18 micron to 0.13 micron eventually got AMD an increase of 600 MHz in production processors, on a similar architecture. That's the fastest Palomino, the 1.73 GHz 2100XP through to the fastest Barton, the 2.33 GHz 3300XP. That's over a 1/3rd increase in clockspeed through process changes. It didn't happen overnight obviously, given that the Thoroughbred A came first, then the Thoroughbred B, then finally the Barton. But it was still quite an increase. 0.13 micron went as fast as 2.4 GHz if you count the Athlon 64s and Opterons, but that was a whole new architecture.

Going from 0.13 micron to 0.09 micron got AMD an increase of 600 MHz - from the Athlon 64 3700/3800/4000 at 2.4 GHz to the Opteron 256/856 at 3 GHz. Not as much this time, but still 25%. Again, it didn't happen immediately, there was the revision D processors (basically Winchester) before we got to the revision E chips (Venice, San Diego, Troy etc).

So using previous form, I expect rev F - the 0.065 micron processors - to get to around 3.75 GHz, in production processors. Given that there has always been overclockers who have managed another couple hundred megahertz out of the chips, I see 4 GHz being plausible. It may take another revision to get to that stage, but I do see it happening. Funnily enough, in talking to some process engineers, we reckoned AMD would get about 3.8 GHz from a die shrink, and that was before we did the quick arithmetic.

In light of that, Intel likely will remain top dog since they have headroom with Conroe. But AMD are at least going to be competitive again, and on performance, not just price. Not that they really care if the relatively few enthusiasts have been lost to Intel, they are still selling everything they can make and that is before counting the Dell sales.

3.75GHz in prduction processors wont happend anytime soon.

90nm on air get to 3100mhz and that is with a 3-4 year old process with like 4-5 revision of cpus and a better socket.
09-09-2006, 09:18 PM
freeloader

I'm just looking to get a 3ghz plus chip on air that consumes less than 100 watts. I'll be happy.
09-09-2006, 09:31 PM
Raybo

Quote:

Originally Posted by ozzimark

i guess you know how well 65nm is gonna clock then? ;)

Thanks for the update.

AMD lost its crown, did anyone else not know this?

:slap:
09-09-2006, 09:37 PM
dinos22

Quote:

Originally Posted by Raybo

Thanks for the update.

what update did ozzimark give besides the ambiguous wink smilie

Show 100 post(s) from this thread on one page