• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

AMD VS Conroe... what AMD is upto

Status
Not open for further replies.

Serge84

Gawd
Joined
Nov 5, 2005
Messages
693
Conroe...
The conroe uses 4mb's of L2, it is a SSE monster. 2 cores, extra 64 bit line. About it. Will be a great multi-tasker. But its really not that impressave. It cheats by using that 4mb's of fast L2 for long processes meaning it seems faster then it really is. But when it runs out of L2 its no more as good as a FX-60. The conroe is great for benching and sprinting. But won't be as good in games as a next gen FX.

Conroe is impressave don't get me wrong. But its also on 65nm's. AMD hasn't wanted to steap that way yet. They really like to keep there lips shut. But they been knowing how to goto 65nm's for along time. Even 45nm's. But it takes then a year to switch to diff processes. They been experamenting for awile. Amazing 90mn tech really stood up to intel for this long AMD has been using. AMD currently is still the best if you don't include conroe. Intel had to totaly change the way they made cpus to out beat AMD for the molment. But like always AMD hides tricks up their sleve.

AMD is experementing on perfecting K8 arcitecture. They managed to perfect 90nm tech on the AM2. 65nm's come later for AMD cpus. Always be patent to wait for the best. Intel's 90nm tech was flimsy and amazingly inefficient. It got better with 65nm's for intel. But they are only 2nd best. Intel always rushes tech out the door cus they are afrade. See how AMD always is level headed, you don't see AMD rushing AM2 out the door. Oviously something doesn't have them shaking in their boots. Not cus they don't have the tech. They just don't really care.

AMD...
Heres what AMD has done to finalize the 90nm arcitecture. It can't get any better then how AMD has made it now. Its even more efficient, they improved speed per clock abit, Raised multiplyer FSB from 200 to 333 and HTT to 1300. They made L2 improvments and are experementing with smaller S cells. They will switch to a faster type of L2 called Z cells and use L3 later. Almost amedeitly Socket F will have 4 cores. The CPU's will be larger and faster then the AM2 types and be using the pad socket types with 1200+ pins. L2 is more dense in these new AM2 chips. Use 940 standard pins. Run a HTT at 2600MT/s. Use DDR2 and DDR3 later as well as other types of fast ram.

Only a 10 to 15% increase in speed. What can you expect from 90nm tech tho? It is the max they can get out of the process. Just see what happens with 65nm's and AMD. Will be a much nicer difference. Anyways the core itself has been made in such a way where its 90nm perfection. They put new types of instructions into it and another 64 bit path. They changed the heatsink bracket itself so it could take much larger heatsinks so they could increase their clock speeds twords more what the newer boards and chips are capable of since its a new socket the limits of clock speed are no longer the same at stock. Besides the fact that AMD makes chips underclocked so they can be overclocked.

AMD chips prob never run at real stock on what they are tested on. Usouly can get 1gb out of them nomatter what in a OC. And 1.5gb in a good day mostly with the semperons. This will be doubled because of the slightly different desine still K8 but with verious improvments to increase performance and tollarances. This means a great deal to OCing. No cap leakage at high speeds and FSB as well as volts being the main factor. Intels are very well known at cap leakage. AMD has made the multiplyer to 333 per x now. making data transfure about 40% faster and raises the max FSB for the cpu's threshhold from around 400 or 500 in a OC or higher. You'll see DFI boards with a FSB multiplyer at around 600 max. Usouly 450fsb is max on current boards like my Tforce6100 or any DFI boards. Tho almost no one can ever brake 400fsb. Rare they can. Well in the extreme forums they like to show off like that. Some just get lucky.

Future...
That socket type is capable of stock speeds of more around 4ghz. We'll see speeds on AMD's line at around 2007 and 5ghz at 2008. Ahh the laws of speed. Doubles every 2 years. Well thats speed wise on how AMD clocks their cpus much differently then intels. But I know AMD is going to have better speed per clock over time so 3ghz in 2 years to AMD is going to be like 8ghz to intel now as in how intel clocks their cpus until . Conroe... that is then intel changed the way they clocked cpus becuase its faster per clock then current intel chips.

Conclusion...
So whats the point of all that. I just said what most know maybe. Just expect more from the master. Thoses who hold out win in the end. Just because intel wins in the sprint doesn't mean they will win when the time comes. Hay hasn't this always been the case? Intel comes out with something that beats amd then amd comes to beat there ass all over again am I wrong no? AMD puts the wool over our eyes for a reason cus they don't want intel to know what they got cooking and its always intresting to see what they come out with. Socket F will be much better then socket AM2 remember. We know even less of the server line then we do of AM2. And not everything has been released. All we seen is test samples of both. And nothing is ever great at launch.

The FX-62 isn't even close to the best cpu we will see. They do have lines of them upto FX-70's ya know in proto type right now. Infact in 65nm's as well but only prototypes, not cus I know. But I just know how devlopment works around factories, it takes a sertent amount of time and teams to develop the tech and make then reasearch and experement on tweaks and such over a period of months on sertent cpus. They have been for the past year in the works. A cpu consept is made 6 months before it even comes out. And in most cases proto types over 3 vers are aready out in the labs on release being experemented and tweeked up. Others more close to release are given to a selected few to be installed and tested under real world experements such as what the cpu can really do on its own. These would be pre production sample prototypes. Almost finals of the product if not the final depending on how good it does, and given to a trusted AMD employee such as one in the development team or a tester. They see if more tweaks need to be done and such. Test its performance and junk and reports back to amd about the test sample then they make final samples to be shown off to the world at a show of some kind. The 1st production modal at a show floor. Development takes years and they been on AM2 S1 and F for awile. Witch is a true assumption. As with intel. We never know whats in the works and what they show on launch is never the badest thing they have. They always have a backup. But intel did seem pritty much in a rush to get their SSE monster out now didn't they to battle AMD... wonder why? Cus intels 65nm stuff didn't live upto the hype. AMD still kicking ass with that old 90nm's now arn't they. Pushed abit too far did AMD? At the last of intels rope perhaps? Ya never know... but one fact remains clear AMD never rushed AM2 out. They took there time way before even the SSE monster was shown by intel. AMD isn't impressed. Well IBM made the tech... intel isn't useing the tech very well.

I mean come on now... look AMD's FX-60 will distroy any intel EE with 65nm tech. AMD's are still currently faster. Thats just the facts. consittering 90nm's agenst 65nm's should make the difference. Well intel had to make a new ark to get them higher finally at a reasonable speed per clock. Intel use to cheat too hard to get that speed too. 4ghz really was more like 2.6ghz or something if ya compared intel to a AMD. Thats silly. Too hot running at 70C and 90C or higher and 150 to 200watts just for the cpu really wasn't my taste. If ya did OCing with a Intel thats what your peak would be in watts. And intels just run until they overheat cus they are programed to to keep the speed up. Oh I just love hateing on the 2nd best now don't I. I guess we all AMD fans love that.

I sure would like both that SSE monster and a new AMD monster. They both don't really impress me, its really not much of a difference. Don't expect a revolution. Only one advancment at a time like this isn't going to make me buy a new system. Or slober over it like a dog in perpetual heat. Its just a better sand box to play with is all. Doesn't mean you'll really see that big of a difference with any 2 machines. maybe in your head. But mixed unmarked boxes with the best of intel and the best of AMD side by side. You'll never tell the difference in a game or anything else. They are the best of the best. They are monsters and what they are. Any system these days is going to be awesome. And when it comes down to it what really matters for games is really the gpu inside not the cpu. But some of us just have to have the best. Go with what your a fan with when it comes out and whats best in the real world. Benches don't matter in the end really Its never enough to even tell unless your told whats in the box. You just can't feel the difference. They are both super fast cpus, with lots of L2, 64 bit, and dual cored. What more could you want?

Well I sure as hell am not going to change my mobo after every new cpu that comes out oh boy intels what fun. >__> Can't intel make one damn chipset that supports all cpus for that generation. 939's supported them all for 3 years. No 775 can support crap for more then 3 months. Just another reason why AMD is better. Intel needs memory controlers in the damn cpu. BTW the cold bug is gone with AMD's new memory controlers. They been perfected as well with total cold resistance. No more limits. Means double the OCing. That maybe the really glory of AMD's. The ultimate overclockers. Well they will have better resistance to everything was one major point to the slight revamp of K8's. They are pritty much K8's in a harder shell. perfect arcitecture and components with v no limit. Thats got to say something to AMD fans. We like to OC. Now we have more to look forward too. I can't wait to see how AM2s and socket F's OC when they come out.

We'll see what really happens on 6/6/6. Cus the FX-62 is evil it be the devil. Ma hahahaha. Intel had better be afrade of that mark cus the FX-62 is very evil with a release date like that. lol That cracked me up. I would be afrade to buy a CPU with a release date like that. :p Wouldn't we all... AMD is evil to even issue a date like that. XDD

Thats all I have to say...
 
At least he broke up the mess into paragraphs. Better than some forum goers.....even if the information is bogus and the grammar is rough to understand.
 
Serge84 said:
We'll see what really happens on 6/6/6.
i'll bite.
you must be hiding under a rock.

first, the date has moved to 5/23

most of that stuff won't happen until 65nm. i'm 99% certain that we don't see any retail quad core opterons on 90nm.. the die will be pretty close the size of the package ;)

i'm almost positive we'll remain on a 200mhz HTT, but we'll see for sure. the fact that all the AM2 previews i have seen thus far show a 200mhz HTT is what makes me doubt the sudden move to 333mhz. even if they did change it, the HTT speed does nothing for performance

your socket F comments.. well socket F is the server socket, kinda like AM2. we don't care about it.

lastly, a 1gb overclock? damn, i wish i could overclock more capacity into my ram :p
 
Serge84 said:
Conroe...
The conroe uses 4mb's of L2, it is a SSE monster. 2 cores, extra 64 bit line. About it. Will be a great multi-tasker. But its really not that impressave. It cheats....

that's all I needed to read.
 
Seriously, 5 GHz in 2008? I wonder what an AMD engineer would say to that...after he was done laughing hysterically, that is.
 
xonik said:
Seriously, 5 GHz in 2008? I wonder what an AMD engineer would say to that...after he was done laughing hysterically, that is.

He would say, "Look, we may be doing a really good job right now, but we aren't God."
 
lol, I won't even try to resist.

That is the biggest pile of bullshit I've ever read in my entire life!
 
Nice stuff! :D As far as I know there is some truth, but I dont want to comment what is and whats not. Im to tired after I read this :eek:

And we all have rights to express opinion!
 
ToastMaster said:
My eyes can only take so much bad grammar. does someone have a cliffnotes version?

  • Conroe is fast in benchmarks - lots of cache is cheating though.
  • You can get a 1-1.5gb overclock on AMD
  • Amd has Z-RAM
  • AMD has FX70
  • 5ghz in 18 months.
 
FreiDOg said:
  • Conroe is fast in benchmarks - lots of cache is cheating though.
  • You can get a 1-1.5gb overclock on AMD
  • Amd has Z-RAM
  • AMD has FX70
  • 5ghz in 18 months.

lol, thanks.

1) Conroe isn't cheating. It's simply powerful.
2) I've seen a 1Ghz overclock on Opty 165, but yeah, that's the fastest I've seen.
3) AMD does have Z-ram, but that won't play a factor for a few years still (if ever).
4) I'd certainly like to see a FX-70.
5) 5 ghz? in 18 months? LOL
 
ToastMaster said:
lol, thanks.

1) Conroe isn't cheating. It's simply powerful.
2) I've seen a 1Ghz overclock on Opty 165, but yeah, that's the fastest I've seen.
3) AMD does have Z-ram, but that won't play a factor for a few years still (if ever).
4) I'd certainly like to see a FX-70.
5) 5 ghz? in 18 months? LOL

I had a 144 that could do 3Gig all day, a 1.2ghz OC. Also my 165 does 2900MHz all day, a 1.1ghz OC.

A 1.5ghz OC would be possible on... a 2800 or 2600+ sempron with a lot of luck, making for a 3100MHz final clockspeed.
 
robberbaron said:
I had a 144 that could do 3Gig all day, a 1.2ghz OC. Also my 165 does 2900MHz all day, a 1.1ghz OC.

A 1.5ghz OC would be possible on... a 2800 or 2600+ sempron with a lot of luck, making for a 3100MHz final clockspeed.


I want your 165 :(
 
ToastMaster said:
2) I've seen a 1Ghz overclock on Opty 165, but yeah, that's the fastest I've seen.
No, not GHz.. "gb."

Thanks for the read serge, but it did get kind of long after a while..
 
Well... I thought it was a fascinating post. I had no idea there were so many ingredients in cheesecake.
 
How is using more cache cheating?

This is just taking advantage of the situation....

More cache is better always, is it not?

what are they suppose to do? have this ultra power chip and but 64k of cache on it?

(You know, I mite as well paraphrase it but.... who would ever need more than 64k of memory?!)

lolerskates

Conroe for the win... I am NOT an intel fan really either, they simply designed a good chip.

So When I am using my conroe in the summer, everytime i turn on my PC I will think to myself, well i am a dirty cheater.... 4MB of cache really is sneaky

sigh.... I am glad I am edumakated
 
USMC2Hard4U said:
How is using more cache cheating?

This is just taking advantage of the situation....

More cache is better always, is it not?

what are they suppose to do? have this ultra power chip and but 64k of cache on it?

(You know, I mite as well paraphrase it but.... who would ever need more than 64k of memory?!)

lolerskates

Conroe for the win... I am NOT an intel fan really either, they simply designed a good chip.

So When I am using my conroe in the summer, everytime i turn on my PC I will think to myself, well i am a dirty cheater.... 4MB of cache really is sneaky

sigh.... I am glad I am edumakated

Only one time can I think of "more cache" not helping every time, and that's the 6xx series 2mb cache. Extra latency FTW!

But Conroe having 4mb cache certainly isnt cheating. It's merely taking advantage of Intel's superior cache design.
 
I wonder where this kid learned grammar...

Anyway, I like AMD as well, but you can't deny the fact that conroe performs better in many cases. I hope AMD has something in development to compete.
 
DarkenReaper57 said:
I wonder where this kid learned grammar...

Anyway, I like AMD as well, but you can't deny the fact that conroe performs better in many cases. I hope AMD has something in development to compete.
I expect them to compete, but nothing that will be a quick fix, K8L is sometime off for the Desktop front, and is designed to counter Woodcrest/Clovertown not Conroe.
 
Do you guys think that anwser to Conroe lies in K8, or do you think it's going to take a new architecture from AMD to compete with Conroe on a level playing field?

I believe they can modify the K8 to get up to Conroe levels of speed (might take a year or two) but I think it will take a new architecture to actually beat Conroe by a good 20-40%
 
By that logic, Opterons and FX-60s also cheat. I mean, look at them, they have double the cache of a normal AMD processor! Like, that's not even fair, Opterons and FX-60s should be removed from all benchmarks and be kicked out of the enthusiast community!! Extra cache...pfft, what a cheesy tactic for menial speed boosts. :p
 
:D

i think someone has it in for AMD.

in all seriousness it was kind of like a fascist post, well with the exception of the grammar and all, but very hitler like.

dont give in.
/end.
 
well, first off - props to any person who can write that much about a chip and not get paid for it. But, I know you didn't get paid for it bc of the grammar/spelling. But, that is petty. I think that you have good intentions, OP, but you are a little misguided.

But, what the hell do we know. We aren't AMD engineers - maybe they do have FX-70s and 5GHZ in 18 months. Maybe, not likely from the situation now - but maybe.

All the people making fun of the gb/ghz thing - get over it - not a big deal.

....meh
 
fuelvolts said:
well, first off - props to any person who can write that much about a chip and not get paid for it. But, I know you didn't get paid for it bc of the grammar/spelling. But, that is petty. I think that you have good intentions, OP, but you are a little misguided.

But, what the hell do we know. We aren't AMD engineers - maybe they do have FX-70s and 5GHZ in 18 months. Maybe, not likely from the situation now - but maybe.

All the people making fun of the gb/ghz thing - get over it - not a big deal.

....meh

I agree with you 100%

But it does bring up an interesting situation. We all pretty much now that Intel will have a decent chip out this year. Something that will take the lead on the desktop.

Now My guess is that AMD is not sitting on there hands. In 18 months where are they going to be? Who knows.... That is a long time out..........
 
Serge84 said:
Conroe...
The conroe uses 4mb's of L2, it is a SSE monster. 2 cores, extra 64 bit line. About it. Will be a great multi-tasker. But its really not that impressave. [ ...]

that is quite an impressive observation. Nonetheless we all know that using more and 'better' L2 cache is not cheating, it is enhancing performance, which is what most of us would like.
Serge84 said:
AMD hasn't wanted to steap that way yet. They really like to keep there lips shut. But they been knowing how to goto 65nm's for along time. Even 45nm's. But it takes then a year to switch to diff processes. They been experamenting for awile.

I am sure that AMD desires to step their process down to 65nm and 45nm. Experimenting with a new process and actually producing a retail product are two very different things. You admit that it takes them a year to swtich to a different process, so it appears to me that Intel is ahead in terms of production process.
Serge84 said:
at some point in here I was bored and realized that you are not a native speaker, so I will cut you some slack with the spelling mistakes.

Bottom line: Just like I do not give Conroe any praise before it is out, I won't give AMD the benefit of the doubt either. I will make my decision on the better product at the point when I will upgrade.
 
drizzt81 said:
I am sure that AMD desires to step their process down to 65nm and 45nm. Experimenting with a new process and actually producing a retail product are two very different things. You admit that it takes them a year to swtich to a different process, so it appears to me that Intel is ahead in terms of production process.


This is a little bit misleading. Nothing terribly wrong with it either. I just want to put one thing out.

Alot of people equate a higher resolution as being the better process, and that just simply is not the case. AMD's 90nm process is performing on par, and in some cases better then Intel's 65nm process. Plus the properties of SSOI allow them to do things in silicon that Intel's process doesnt.

The only advantage that I can think of is that Intel can produce higher resolutions...
 
Where do you get this stuff? It's too good to be made up :p

In all seriousness, you sound like a real AMD fan. Open your eyes for a second, and stop making excuses for AMD. You make a lot of assumptions, and throw some bad logic in the mix.

This paragraph made me laugh the most:

I mean come on now... look AMD's FX-60 will distroy any intel EE with 65nm tech. AMD's are still currently faster. Thats just the facts. consittering 90nm's agenst 65nm's should make the difference. Well intel had to make a new ark to get them higher finally at a reasonable speed per clock. Intel use to cheat too hard to get that speed too. 4ghz really was more like 2.6ghz or something if ya compared intel to a AMD. Thats silly. Too hot running at 70C and 90C or higher and 150 to 200watts just for the cpu really wasn't my taste. If ya did OCing with a Intel thats what your peak would be in watts. And intels just run until they overheat cus they are programed to to keep the speed up. Oh I just love hateing on the 2nd best now don't I. I guess we all AMD fans love that.
 
Status
Not open for further replies.
Back
Top