Conroe owns AM2

[ColdFlo]

I know its sad I wish AMD could stay ahead but when I saw the pentium m benches that told me that intel had a way back in. The pentium m benches and ocs I have seen are pretty amazing and now I've found AM2 benches and I'm sorry folks but I think the AMD ride may be exiting as we speak.

http://www.hkepc.com/hwdb/am2-4800-1.htm

Sorry for the bad news sucked the wind out of me too I was really looking forward to the upgrade now I'm stuck with a system that cant play Quake 4 very well and a Conroe release in July. Looks like ill be sucking rockets for the forseeable future. Noooooooooooooooo.

[chew*]

I wouldn't count AMD out yet..........They may be pulling a rope-a-dope on intel.......What that means is they may let intel think they have a comfortable lead.........

What gives me this impression you say? It all boils down to stepping. 0543 MPMW is a very very early batch of AM2.........I would take those results with a grain od salt for now.

Not to take the spotlight off of intel, they do have some serious stuff coming on the horizon. It's a big improvement for them thats for sure. I wouldn't count out the guys in green just yet though. Wait till its all been officialy released then draw your conclusions.......

[situman]

I wouldn't count AMD out yet..........They may be pulling a rope-a-dope on intel.......What that means is they may let intel think they have a comfortable lead.........

What gives me this impression you say? It all boils down to stepping. 0543 MPMW is a very very early batch of AM2.........I would take those results with a grain od salt for now.

Not to take the spotlight off of intel, they do have some serious stuff coming on the horizon. It's a big improvement for them thats for sure. I wouldn't count out the guys in green just yet though. Wait till its all been officialy released then draw your conclusions.......

AMD is not smart enough to play rope-a-dope. They are more about reaction than prevention. Socket 754 ring the bell? Even if AMD is able to squeeze another 200 or so MHZ from 90nm, it would still fall behind Intel. If AMD is trying to win via increasing speed, then it just assumed Intel's former position. Amd lost this round no doubt and they got caught with their pants down this time. It would be pointless to redesign a chip based on 90nm and make it a very short lived chip. AMD simply does not have that kind of resources.

The bright side is that they've seen what Conroe can do and still have about a year to tweak their 65nm product. Let's hope they grow a brain by then and hopefully won't release another socket cause ooopps socket AM2 can't support 65nm chips.

[Absolute_0]

It's not like AMD has been laying dormant all these years. AMD 64 is getting old... they are working on new stuff ;)

[lapdog]

It's going to be next year before AMD has a chance with 65nm, and Intel has 45nm set latter next year. I hope AMD has a new design in the works or it may be years before they get back in the game. K8 is great, but Conroe is better. Everyone's blasting the FSB, but it must be pipline stalls slowing down the prescott.

[chew*]

Don't get me wrong. I'm always up for a change......I have used nvidia,ati,amd intel. I just think it isn't over till the fat lady sings. I'll wait it out. 3.0 gig dualcore should be able to tide me over for awhile.

[Richteralan]

I really don't know why everyone get excited about Conroe.

Is it released yet?
Can you buy it yet?

Unfortunately both anwers are NO.

Yes after reading the preliminary performance numbers I would say Conroe is good. But that's it. I'll hold my judgement AFTER I can see some real product in mass quantity.

[Richteralan]

I know its sad I wish AMD could stay ahead but when I saw the pentium m benches that told me that intel had a way back in. The pentium m benches and ocs I have seen are pretty amazing and now I've found AM2 benches and I'm sorry folks but I think the AMD ride may be exiting as we speak.

http://www.hkepc.com/hwdb/am2-4800-1.htm

Sorry for the bad news sucked the wind out of me too I was really looking forward to the upgrade now I'm stuck with a system that cant play Quake 4 very well and a Conroe release in July. Looks like ill be sucking rockets for the forseeable future. Noooooooooooooooo.

Yes It's sad that you post this stuff, too.

[Repoman]

Yup, I plan on keeping this rig until at least next spring, hopefully longer! CPU doesn't have much to do in current games anyway

But yeah, I do think my next rig will be a Conroe, no denying that. We already know how Yonah does, and obviously conroe will be even better than that, so..

[G4h4o8s6T]

I read here in another thread about AM2 eventually moving onto the K8L architechture that will be used for socket F Opterons.......double the FPU units, and SSE improvements.

I dont know much about this, just what I have read here......it may not beat out Conroe, but dont you think its a possibility it will atleast be at the same level???? It seems that the initial release of AM2 wont have a chance against Conroe, but do you really think AMD is just gonna sit around and let Intel blow them out of the water.........I think not

[pcdoc1]

You know the guys over in the Intel forum have got to be loving this… They’ve been sucking hind tit for years now as NetBUST has always run 500mph while going nowhere. Finally the Intel marketing guys lose control and the techs are back in charge, they roll out what appears to be a good/decent product and then all the “LOYAL” AMD guys just roll over…

I’m with Chew*, this party is far from over. I’m not sure this party has really even began… But one thing for sure, you can thank AMD for bringing Intel back to reality… It wasn’t that many years ago that anything from Intel that even remotely resembled performance started at a $1000.00… :)

[topaimz]

IMO, AM2 socket is a waste :p:

They should keep making s939 processors till they get something to own conroe's (65nm) arse with...
Because it is pretty obvious that with intel bent on
bringing AMD's popularity down, AM2 is only gonna last a while till AMD makes
their own smaller, less power consuming yet better performing line of processors.

Is it me, or does AM2 remind you of 7900s... not much gain in performance, and
acting like a transition bridge/cover for the time being (to hold the market), until the actual competitor is released? (can't wait for G80s specs :D ) :p:

[ReD.SkY]

wait, is AM2 65nm?

[nfm]

Intel finally went smart and realized that pipelines hurt the performance of netburst architecture at lower frequencies. Take a look at older northwood cores, they had I believe 21 pipes compared to Prescott's 31. Performance of same model of both cores were almost the same, but actually northwood performed better than prescott. You can guys see what took to Intel engineers to minimize lost performance in prescott. If prescott would had have 21 pipes in the beginning as northood did, then we would see some good improvements. Prescotts were designed for high mhz but didn't Intel know that it is impossible to run 5Ghz+ on copper heatsink with current nanotechnology and current leakage? I see no point in prescotts, but we've seen crazy japan folks running them at 7Ghz+, and they really fly at those speeds and pipelines are fully utilized.

Right now what I can tell you is that intel leaped and is in the lead. With 14 pipes and technology descended form Pentium M we see a real potential and extremely good efficiency. Keep in mind that conroe runs on high latenicies as all Intel based chips do. Intel still has a lot of headroom, they can include on die memory controller to minimize latencies and gain easy performance.

I wonder what other changes took place, as for right now, IBM G5 has the worlds most advanced Execution Units and all microprocessor's internal processor structure. But somehow G5 fall way back behind AMD and Intel chips in real world performance.

I'm also excited about imroved instructions, SSE4 should also be included. Those simple set of insctructions can really improve performance, why do twice as much math when specific intructions can help you do to it at once? ;)

Just some of my thoughts from top of my head, good thing is that Intel leaps in right direction and they have a lot of silicon available...

[ZhaoYun]

In moments like these, the only thing we can count on is patience. :wiggle:

[nn_step]

Honestly you all have absolutely NO idea what is going on..

[pcdoc1]

Why don't you enlighten us.........

EDIT: 24.86 Average daily post of nonsense starting 9-05

[amrgb]

I wouldn't count AMD out yet..........They may be pulling a rope-a-dope on intel.......What that means is they may let intel think they have a comfortable lead.........

What gives me this impression you say? It all boils down to stepping. 0543 MPMW is a very very early batch of AM2.........I would take those results with a grain od salt for now.

Not to take the spotlight off of intel, they do have some serious stuff coming on the horizon. It's a big improvement for them thats for sure. I wouldn't count out the guys in green just yet though. Wait till its all been officialy released then draw your conclusions.......

If they are doing what you say, then it only shows how poorly they do business. There's nothing to hide. Intel put the cards on the table for the next round. What can Intel do in 6 months to react to a Super Charged AM2? Nothing.

AMD is :wasntme: like Intel was for long time. It's the role of the player in disadvantage, to be quiet. Remember Intel ignoring AMD challenges to run benchmarks etc? Now it's time AMD ignore Conroe like it doesn't exist, and hopefully concentrate efforts in developing something to fight Intel, which AM2 isn't (it seems, I would be glad to swallow this sentence after AM2 launch). It's like in the automobile industry. AM2 is just a mere restyling, not a new model.

[nn_step]

Why don't you enlighten us.........
Sure just wait two months and I'll show you...

[r3w4]

You guys need to stop gossiping about little information based on products that won't be out for a while. This is exactly what these previews are supposed to do. "Oh crap I better not buy that [insert AMD processor] because this one benchmark says Intel will have something good in 6 months." Lots of things can happen in half a year.

[topaimz]

wait, is AM2 65nm?

Nope. As stated above, AM2 is just a 'restyling' of the s939.

Only thing that holds a great significance with the launch of the AM2,
to my knowledge, is that it can utilize DDR2, nothing more, nothing less...

[pcdoc1]

Sure just wait two months and I'll show you...Why wait, you haven't had anything to show us since you got here except...
24.86 Average daily post of nonsense starting 9-05

[nn_step]

Nope. As stated above, AM2 is just a 'restyling' of the s939.

Only thing that holds a great significance with the launch of the AM2,
to my knowledge, is that it can utilize DDR2, nothing more, nothing less...
:stick: You honestly think AMD spent the last two years designing AM2 and the only thing they did different was a new Memory controller :stick:

[topaimz]

:stick: You honestly think AMD spent the last two years designing AM2 and the only thing they did different was a new Memory controller :stick:

well, thats all I can see of this moment, and stop poking me! :stick:
Im sensitive dang it! I get agitated when I hear my fans make around ~30dBA
of sound! :stick:

Find something that would show us that AM2 will be more then what I expect:stick:
then I'll stop :stick: you, and AMD's attempts

Cuz from what I can see, the initial launch of AM2 processors are just
remake models of the x2s and maybe one new FX dual core :p:

P.S.
I think AMD should make more s939 processors :p:
People like me who bought 4400+ and above feel cheated :(
cuz basically, the 4400+ -> FX60 is same core :(

[fhpchris]

It wasn’t that many years ago that anything from Intel that even remotely resembled performance started at a $1000.00… :)

The 2.4C kicked the crap out of everything Amd had untill sledgehammer dropped.

Amd is not stupid, they have something waiting, and they will use the Hypertransport/integrated memmory controller to their advantage.

Lets not forget that 7 gigaflop co processor. Tides can turn fast.

[pcdoc1]

The 2.4C kicked the crap out of everything Amd had untill sledgehammer dropped.

Amd is not stupid, they have something waiting, and they will use the Hypertransport/integrated memmory controller to their advantage.

Lets not forget that 7 gigaflop co processor. Tides can turn fast.Yep, I had one at 3.6 and it was kool at the time....... As I said earlier this party hasn't even begun let alone be over... We'll see soon enough.

Hell, we might even finally see some pics/results from nn_step: Oh wait, I'll bet he's under NDA! :slap:

[CRUSH-ER]

I for one can't wait for Conroe. I've been missing Intel....almost bought a P-M but decided not to. I love AMD too. Competition is a good thing, don't play sides. :)

[r3w4]

Find something that would show us that AM2 will be more then what I expect:stick:
then I'll stop :stick: you, and AMD's attempts

Cuz from what I can see, the initial launch of AM2 processors are just
remake models of the x2s and maybe one new FX dual core :p:
I wouldn't rely on who gives information first and utilizes it best. You're just fulfilling marketing's purpose if you do that.

[Dumo]

We can all wait for both platforms and go from there:) ...A few benches don't mean a thing.
Personally, I want to try both and stick with the best...limited to benchs....don;t really care about other else:toast:
Basically we have to choose any combination of Nvidia, Ati, Intel and Amd....or get all.

[biohead]

i like tea and biscuits.

[Dumo]

13X270 or use divider.....Hope is not XXX2E or 3E:)

[topaimz]

lol, I see an alienhead! :stick:

[don_vercetti]

Subscribed...i can see this taking an interesting turn.

[IluvIntel]

You guys need to stop gossiping about little information based on products that won't be out for a while. This is exactly what these previews are supposed to do. "Oh crap I better not buy that [insert AMD processor] because this one benchmark says Intel will have something good in 6 months." Lots of things can happen in half a year.


Agreed, You guys have no idea what its like benching with 65nm P4's - I just got off air only just shy of 5.0ghz, solid benchmark stable with minimal Vcore - how's "auto" volts in bios for 50% overclock sound ?
Then only 5% more for 4.98Ghz. :D :D :D
If this is an example of 65nm fabrication efficiency with "old" netburst architecture, then Conroe is going to king of desktop cpus for sure !

[chew*]

The 2.4C kicked the crap out of everything Amd had untill sledgehammer dropped.

Amd is not stupid, they have something waiting, and they will use the Hypertransport/integrated memmory controller to their advantage.

Lets not forget that 7 gigaflop co processor. Tides can turn fast.

Hyper transport speed is over rated.....I can run at 1500 HT or even 1600 and get a 4 second gain in PI 8m which amounts to a hill of beans in real world performance. The memory controller already exists and although they can tweak it it won't net much more performance as its already been there. Amd is going to have to come up with something besides this and clock speeds IMHO. Not sure what but it's going to be, I'm just sure it will be necessary to be competitive...

[Blacklash]

Sad? I don't think so. The more competition the better.

I will be buying the best tech, and could not care less who makes it. I respect AMD and have enjoyed their products, I don't own stock in their corporation however.

I'd wait until we see full reviews from up to ten or more hardware sites before we start talking about a clear winner. As always never ever go off a single review when buying hardware.

[vengance_01]

AMD is not smart enough to play rope-a-dope. They are more about reaction than prevention. Socket 754 ring the bell? Even if AMD is able to squeeze another 200 or so MHZ from 90nm, it would still fall behind Intel. If AMD is trying to win via increasing speed, then it just assumed Intel's former position. Amd lost this round no doubt and they got caught with their pants down this time. It would be pointless to redesign a chip based on 90nm and make it a very short lived chip. AMD simply does not have that kind of resources.

The bright side is that they've seen what Conroe can do and still have about a year to tweak their 65nm product. Let's hope they grow a brain by then and hopefully won't release another socket cause ooopps socket AM2 can't support 65nm chips. Yea you have to remember when the K8 chip was released. Its on-die mem controller was a pretty big thing back in the day. I am sure AMD has something up there sleeve. But have we seen this trend before?

AMD and intel have been trading blows for quite some time. I still have to agree Intel has one this round and its a solid chip by Intel. But look at the market share AMD has gained with the A64. It was a great product.

[don_vercetti]

won this round????
one: i think its wrong to call them rounds.
two: If there was ever a round, it was K8 vs. Prescott. And i think we all know who won THAT round. The 'next round' is Conroe vs. amd 65nm products.

[gundamit]

If AMD did have something up their sleeve they would have leaked it by now if only to kill some of the enthusiasm for Conroe. To AMD's credit no FUD campaign has emerged. Early AM2 results seem to point to little if any performance gain. The link in the original post seem to take memory performance out of the mix leaving just a comparison of the CPUs. Even if AM2 brings more bandwidth we all know that won't make a dramatic difference.

If my budget allows (or if Intel misses its release dates) I may go for both just to play with them. AM2 should be out before Conroe so those who only want to pull the trigger once will have plenty of time to see what AM2 brings to the table before the Conroe option is available.

There is more then enough evidence at this point to get a little excited by Conroe. I hope Intel isn't just teasing on the the release dates or pricing just to cut into AMD's current sales.

Nothing but good news here. More competition, more choices as the performance war heats up. :D

[Kaiser_Sose]

am2 reminds me of the bogus worthless Intel move to 915 chipsets ... they both are unnesessary and worthless

one problem with these new conroe systems is going to be the price ... the price of the conroe board alone will be close to the combined price of an and cpu and board

[dball3]

the only good thing i have read about am2 is there's a good possibility you will be able to slap some ddr3 into it when that time comes.:D

[biohead]

the only good thing i have read about am2 is there's a good possibility you will be able to slap some ddr3 into it when that time comes.:D
it all depends on the law suit..

[r3w4]

If AMD did have something up their sleeve they would have leaked it by now if only to kill some of the enthusiasm for Conroe. To AMD's credit no FUD campaign has emerged. Early AM2 results seem to point to little if any performance gain. The link in the original post seem to take memory performance out of the mix leaving just a comparison of the CPUs. Even if AM2 brings more bandwidth we all know that won't make a dramatic difference.

If my budget allows (or if Intel misses its release dates) I may go for both just to play with them. AM2 should be out before Conroe so those who only want to pull the trigger once will have plenty of time to see what AM2 brings to the table before the Conroe option is available.

There is more then enough evidence at this point to get a little excited by Conroe. I hope Intel isn't just teasing on the the release dates or pricing just to cut into AMD's current sales.

Nothing but good news here. More competition, more choices as the performance war heats up. :D
I wouldn't assume anything. AMD is a smaller company than Intel and doesn't have time to waste playing "look how big my :banana::banana::banana::banana: is now, if you measure it this way" games. It's sad how Intel can make a shift in crowds by releasing some small info on future products.

[Getttosmurf]

il probably do what i always do wait untill someone here post som benchmarks and then wait another 3 months and buy when the gear lands in sweden ;)

[irev210]

I think many forget about cost. Intel is HUGE supplier to OEM, like Dell. They are looking to make CHEAP processors.

The moves to 65nm and 45nm is a cost move, performance is just a bonus.

AMD will fight back, but the battle will be much more balanced.


Regardless, nothing will be as clear cut as it was last year with prescott vs amd 64.

This year, it will be much closer, and much more fun!

[muzz]

I look at it like this myself....

Amd has basically owned Intel for a bit, Intel smartened up and finally came up with a solid product.
Thats good for ALL puter fans IMO, the more competition, the more advance in technology.
The Conroe chip looks great in the benches I've seen so far, but there are no Conroes available that I know of.

I doubt AMD has been sitting still watching this doing nothing though, so IF the Conroe is all that ,AND a smelly thong, then we should be seeing something from the AMD camp to compete in fairly short order.
If NOT, then Intel will own THIS round.... nothing more really.

I PERSONALLY am not expecting big things from AM2, but I could be wrong.....

I like compettition myself, I dunno about most folks here.

m

[i found nemo]

well, i've said once i'll say again. good for intel. i hope this glorifies so that when amd kicks ass every one will shut up :)

[situman]

You know the guys over in the Intel forum have got to be loving this… They’ve been sucking hind tit for years now as NetBUST has always run 500mph while going nowhere. Finally the Intel marketing guys lose control and the techs are back in charge, they roll out what appears to be a good/decent product and then all the “LOYAL” AMD guys just roll over…

I’m with Chew*, this party is far from over. I’m not sure this party has really even began… But one thing for sure, you can thank AMD for bringing Intel back to reality… It wasn’t that many years ago that anything from Intel that even remotely resembled performance started at a $1000.00… :)

We're not rolling over. Some of us are not "AMDroids." As of right now AMD has nothing to compete with Conroe. Besides, I for one have no loyalty towards either brand. Whoever puts out a competitive product gets my money.

[Orangeman]

Think of it this way...FX-60 will never be beaten by a Conroe. FX-62 will (probably) beat FX-60. FX-62 will be beaten by a Conroe. Conroe will be beaten by (TBD)...

[arisythila]

I think AMD is just seeing who the :banana::banana::banana::banana::banana:es are... one they find out, they will release something better than conroe... Then all the :banana::banana::banana::banana::banana:es that went to Intel, will have to jump back onto the AMD bandwagon.

[muzz]

" As of right now AMD has nothing to compete with Conroe.

Sorry, BUT if ya wanna be realistic it should read...

As of RIGHT NOW, Intel has NOTHING ON THE MARKET that beats the Amd FX-60.

Just trying to keep things current and in perspective.

That COULD change next week, but for now.....

[situman]

ur right.

[Charles Wirth]

Not true, Yonah has been on the market since early January.

Watch the Yonah competition thread, I dont think any AMD world record will be left when we are done.

It takes a dual opty 185 2.6Ghz (dual FX 60) to beat the Yonah T2600 2.16Ghz at Rosetta ;) Thats AMD 4 cores vs 2 Intel.

Not to get too far ahead, a 2Ghz Merom slaps down a screaming 25s pi (at stock speed)

[nn_step]

Not true, Yonah has been on the market since early January.

Watch the Yonah competition thread, I dont think any AMD world record will be left when we are done.

It takes a dual opty 185 2.6Ghz (dual FX 60) to beat the Yonah T2600 2.16Ghz at Rosetta ;) Thats AMD 4 cores vs 2 Intel.

Not to get too far ahead, a 2Ghz Merom slaps down a screaming 25s pi (at stock speed)
Don't you mean Opty 285?, because the 100 series doesn't have the extra HT links required for SMP

[muzz]

Not true, Yonah has been on the market since early January.

Watch the Yonah competition thread, I dont think any AMD world record will be left when we are done.

It takes a dual opty 185 2.6Ghz (dual FX 60) to beat the Yonah T2600 2.16Ghz at Rosetta ;) Thats AMD 4 cores vs 2 Intel.

Not to get too far ahead, a 2Ghz Merom slaps down a screaming 25s pi (at stock speed)

Are we ONLY talking about Rosetta, or are we also discussing real world/ gaming performance as well?

Lets face it, although this IS Xtreme, there are also OTHER uses of said eqpt, gaming being pretty important to high end rig users..

m

[topaimz]

outta the topic of discussion but I gotta ask.

Do the moderators of this forum have a cat fetish :stick: :cat: :lol2:

Various moderators and users seem to use cat avatars/pics :D

[Orangeman]

The whole internet is infected with cat avatars...

[topaimz]

why not dogs? or even the occasional medium sluppy cup with a pink straw, short hands, eyes and a mouth floating in a pool?

[szukalski]

Cats fit into the whole Xtreme theme.. stay at home and spend endless hours tweaking.
A bit hard to do that when you're walking a dog ;)

[nn_step]

Cats fit into the whole Xtreme theme.. stay at home and spend endless hours tweaking.
A bit hard to do that when you're walking a dog ;)
Damn good point

[Haltech]

Rename this thread..

Xtreme Playground

[amrgb]

I think AMD is just seeing who the :banana::banana::banana::banana::banana:es are... one they find out, they will release something better than conroe... Then all the :banana::banana::banana::banana::banana:es that went to Intel, will have to jump back onto the AMD bandwagon.

Yeah right. Like technical advance in cpu's can be made in months. Do you noticed how much time it took Intel to come up with something to be competitive again? And AMD resources are nothing compared to Intel's. I don't think Intel enjoyed this time with AMD on top. At the end of the day it may have made little difference in Intel's pockets. But no giant likes to be overtaken by a smaller player.

AMD got lucky and caught Intel sleeping with that P4 disaster. Now Intel woke up and it's time for the real fight. Place your bets gentlemen. And may the best win... our money.

[muzz]

Luck? I doubt it.

[r3w4]

AMD got lucky and caught Intel sleeping with that P4 disaster. Now Intel woke up and it's time for the real fight. Place your bets gentlemen. And may the best win... our money.
You don't "get lucky" with things like this. Intel screwed up and they know it. They're finally busting out a good processor and their best marketing.

[Kris]

I don't care who is on top... as long as I have their parts in my rig! We gotta remember that we will be the ones who benifit the most from all of this conroe v am2 hype with BETTER cpu's in the end. My 02 cents

[computerpro3]

Why don't you enlighten us.........

EDIT: 24.86 Average daily post of nonsense starting 9-05


You noticed that too?

:toast:

[M.Beier]

Not true, Yonah has been on the market since early January.

Watch the Yonah competition thread, I dont think any AMD world record will be left when we are done.

It takes a dual opty 185 2.6Ghz (dual FX 60) to beat the Yonah T2600 2.16Ghz at Rosetta ;) Thats AMD 4 cores vs 2 Intel.

Not to get too far ahead, a 2Ghz Merom slaps down a screaming 25s pi (at stock speed)

Nevermind rosetta, it'll knock off AMD at their homefield, 3Dmark2001SE...

[GMX]

Why don't you enlighten us.........

EDIT: 24.86 Average daily post of nonsense starting 9-05

LOL this thread is funny. Honestly nothing caught my attention and made me laugh a bit except PCdoc's comments on the famous spammer nn_step.

[Super strokey]

Why don't you enlighten us.........

EDIT: 24.86 Average daily post of nonsense starting 9-05

isnt august the 8th month of the year?

[zir_blazer]

Conroe is still six months away and the Benchmarks were done by Intel assembled setups, and because we can't replicate in no way the tests, we must see a complete, full third party review to make conclusions. We know also than the first Socket M2 Processors will bring the same performance than what we got now, but we also know that AMD wouldn't stand still and that they should have something new to show too. The K9 Architecture is missing in action and we don't know if AMD actually did finished it but keep the secret. And we didn't saw neither 65nm K8s for power consumption and overclockeability purposes.

@ColdFlo
I didn't know than the Quake 2 star in xquake and tastyspleen servers was a computer enthusiast too. Long time that I don't see you...

@biohead
What the hell is that Screenshot and why no one commented about it? It is photochopped or you will leak some more data? :D

[hawkeyefan]

maybe I'm dense...but,

where does IBM fit into all this? You've got to think the engineers at IBM and AMD have some kind of back-and-forth going on. Does anybody know whether the two have joint projects, i.e. architecture, process engineering, etc.?

[nn_step]

maybe I'm dense...but,

where does IBM fit into all this? You've got to think the engineers at IBM and AMD have some kind of back-and-forth going on. Does anybody know whether the two have joint projects, i.e. architecture, process engineering, etc.?
They have many shared projects

[hawkeyefan]

They have many shared projects

muchas gracias...assumed as much

[ceevee]

I keep hearing over and over as regards AMD:

"AMD wouldn't stand still" would they??

"AMD haven't been doing nothing" have they??

"AMD certainly has something planned" maybe they don't??

Could AMD have gotten complacent this time? Seems like if there was more to AM2 we would have heard about it already. Maybe Intel will take back the lead for a while.



By the way ES Oblivion is amazing at 1920x1200 :D

[chew*]

They have many shared projects

What do you want to bet AMD helped IBM out with the xbox cpu :)

[nn_step]

What do you want to bet AMD helped IBM out with the xbox cpu :)
I honestly Have no freaking Idea about the Xbox...

[chew*]

Its a triple core IBM chip :)

Now back to the subject at hand. There are a few things that AMD can do to improve performance of AM2. One would be to once again boost the memory controller from 128 bit to 256bit. Other things that have been talked about is increased L3 cache. I'm not sure what that would do for performance except to eliminate high latencies from causing a perfromance hit. But once again there a lot of things that can be done. We will just have to wait and see :)

[Dumo]

Sorry, BUT if ya wanna be realistic it should read...

As of RIGHT NOW, Intel has NOTHING ON THE MARKET that beats the Amd FX-60.

Just trying to keep things current and in perspective.

That COULD change next week, but for now.....Core duo beat it once in '05:)

[nn_step]

Core duo beat it once in '05:)
What exactly do you mean by just once?

[Dumo]

I did comparo:) ...http://www.xtremesystems.org/forums/showthread.php?t=90743

More to come...soon.

[zir_blazer]

Its a triple core IBM chip :)

Now back to the subject at hand. There are a few things that AMD can do to improve performance of AM2. One would be to once again boost the memory controller from 128 bit to 256bit. Other things that have been talked about is increased L3 cache. I'm not sure what that would do for performance except to eliminate high latencies from causing a perfromance hit. But once again there a lot of things that can be done. We will just have to wait and see :)
And how on the earth would AMD include a 256 Bits Memory Controller for Quad Channel on a Socket M2 Processor if the Pins for doing so aren't there? That would requiere yet another Socket change. The Memory Bandwidth that Quad Channel gives will be as useless as what DDR-II gives, with the exeption than you plan to support a Quad Core Processor with the existing platform... That is more than a possibility with Socket F (That should actually do something good with all the extra Pins that it got), but not with the Desktop Socket M2. Besides, it would requiere Motherboards with a much more complex designs and more PCB layers, increasing the manufacturing cost significantly for some extra performance that isn't worth the extra cost.
Something similar happens with the Cache L3. We know than the Opterons 8xx can go up to 8-Way with a great scalability even when being only assited by 1 MB Cache L2 per Core, not huge amounts of Cache L3 like the Xeons MP. Why would you want to add Cache L3 to an architecture that doesn't heavily relies on Cache for its performance increases? That would significantly increase the Core size, and that mean less Cores per Wafer and less yields too, increasing more manufacturing cost. And all that for what? Slighty more performance? Again, no thanks.

[ahmad]

What do you want to bet AMD helped IBM out with the xbox cpu :)

I strongly doubt there is much AMD could teach IBM that they don't already know... if it wasnt for IBM, AMD wouldnt have been able to do 2 things:

- Go 64bit
- Move to 90nm process

IBM is still a world leader in hardware.

[etekberg]

I think you guys are missing the main point in the debate...not that I have any great insight or anything, but IMO the real battle is in manufacturing. Who cares what great speed demons you can design, it's what you can manufacture and at what cost.

Intel has ALWAYS owned AMD in manufacturing. So now that Intel has removed it's head from it's ass (hopefully), my money is on Intel.

It's analogous to the x1900 vs 7900 debates. Who cares who slightly wins in performance? The 7900 is half the die size! ATI can't afford to compete!

[pcdoc1]

ATI can't afford to compete! Huh? they currently beat Nvidia pricing at every price point.... What planet are you from?

[zabomb4163]

Huh? they currently beat Nvidia pricing at every price point.... What planet are you from?

http://finance.yahoo.com/q/ks?s=ATYT

"Profit Margin (ttm): -0.69%"


http://finance.yahoo.com/q/ks?s=NVDA

Profit Margin (ttm): 12.74%

[pcdoc1]

http://finance.yahoo.com/q/ks?s=ATYT

"Profit Margin (ttm): -0.69%"


http://finance.yahoo.com/q/ks?s=NVDA

Profit Margin (ttm): 12.74%Good info but my point wasn't manufacturers profitability, when ATI wants/needs to improve there bottom line they’ll move to a different process.

Nobody is walking into Best Buy or calling Dell and basing their purchase decisions on die size or process… :)

[onewingedangel]

its unlikely that amd will radically change the core designs, the move to 65nm will boost their clockspeeds enough to be competitive for now - after all how much more headroom will conroe have already being on a 65nm process? Theres also SiGe to consider, which last I heard could be implemented on either 90 or 65nm process nodes. Even if its only a small batch for 90nm (like the strained silicon 130nm fx55's) SiGe could give enough of a boost to keep Amd competitive until 65nm. At which point the scalability of conroe will come into play on whether it will keep up with faster amd chips (if AMd can clock their 65nm chips high enough). Of course this can only continue for so long, before AMD must change things up a bit, but at least until the introduction of CSI, the scalability will keep AMD the leader in the 4 and 8 way market - so theres not going to be a sudden need to make profits on the desktop, so shifting the focus to cheap desktop chips (semprons) and the 2 and 8 series opterons seems the logical move for AMD.

But we al know whats going to happen, the performance crown is going to move back and forth between AMD and intel, with intel gaining the lead with every new process, and amd catching up and overtaking as they migrate to their new process.

Another factor to consider is that a 65nm a64 with the same cache as the 90nm processors would be considerably smaller than a 4m cached conroe, and probably smaller than the 2m cached version. So AMD don't have to fight it out on the high end, and could make tidy profits in the mainstream market.

[zabomb4163]

Good info but my point wasn't manufacturers profitability, when ATI wants/needs to improve there bottom line they’ll move to a different process.

Nobody is walking into Best Buy or calling Dell and basing their purchase decisions on die size or process… :)

well his point was
"ATI can't afford to compete!". ATI may have a slightly better chip but they are bleeding money. like it or not money plays a very large part of how well a company can compete in the long run.

[r3w4]

I think you guys are missing the main point in the debate...not that I have any great insight or anything, but IMO the real battle is in manufacturing. Who cares what great speed demons you can design, it's what you can manufacture and at what cost.

Intel has ALWAYS owned AMD in manufacturing. So now that Intel has removed it's head from it's ass (hopefully), my money is on Intel.

It's analogous to the x1900 vs 7900 debates. Who cares who slightly wins in performance? The 7900 is half the die size! ATI can't afford to compete!
You have to lose some to gain some. The G7xs have been in production for a much longer time than the R5xxs. If AMD doesn't have a golden bunny to pull out of its hat, they will have to suffer low prices and concentrate on their next project. I just can't believe how cheap those Conroes are going to be.

[pcdoc1]

well his point was
"ATI can't afford to compete!". ATI may have a slightly better chip but they are bleeding money. like it or not money plays a very large part of how well a company can compete in the long run.No argument from me, the point was that the statement "ATI can't afford to compete!" is not what is being seen in the retail market. They are competing quite well. I'm doubtful that ATI management needs any advise from us... With that said time will tell the story. :)