AM2 vs. Conroe - Am I the only one?

Status
Not open for further replies.
D

dr_dirtnap

Limp Gawd
Joined
Aug 4, 2004
Messages
339
The comparision I make between AM2 and Conroe is this:

AMD is in the position that intel was going from socket 478 to socket T
1. Different Memory Support
2. Basically same core
3. No major architectural changes

Intel is in the position that AMD was going from the XP era to the A64 era
1. Same Memory Support
2. Very different core
3. Major architectural changes

For some reason, that is what sticks in my mind. I am less excited about AM2 than I am about Conroe. Conroe is basically new from the ground up, where the AM2 is just another revision of the same thing. We know what to expect with AM2-- simliar performance compared to the s939 generation, we have nothing to compare to Conroe. I was very excited with the release of the Athlon 64, and I am just as excited anticipating Conroe.

Mod Edit: Sorry to say your thread was closed for flaming, due to other posters along the way and toward the end. Papa-Ming
 
I was thinking of building an AM2 rig down the road, but after seeing those conroe benches, I know which route i'll be taking. I'd still like to mess with AM2 though.
 
cant wait for conroe to come out. i was going to grab a athlon 64 x2 but looking at those conroe numbers makes me want one really really bad. only downside is, that i wont be upgrading to 2gb of ram untill conroe comes out (no point buying 2gb of ddr1 ram when conroe uses ddr2)
 
dr_dirtnap said:
The comparision I make between AM2 and Conroe is this:

AMD is in the position that intel was going from socket 478 to socket T
1. Different Memory Support
2. Basically same core
3. No major architectural changes

Intel is in the position that AMD was going from the XP era to the A64 era
1. Same Memory Support
2. Very different core
3. Major architectural changes

For some reason, that is what sticks in my mind. I am less excited about AM2 than I am about Conroe. Conroe is basically new from the ground up, where the AM2 is just another revision of the same thing. We know what to expect with AM2-- simliar performance compared to the s939 generation, we have nothing to compare to Conroe. I was very excited with the release of the Athlon 64, and I am just as excited anticipating Conroe.
Click to expand...

Same here.
 
Finally, I'm not the only one who sees it....

I don't remember there being such a fuss raised with the release of the A64's...yet everyone is suddenly confident that AMD will pull something out at the last minute.

I think it's because all these people who suddenly jumped on the computing world bandwagon also jumped in during the AMD era. Some of these people are also the same type of people who pretend they know what they're doing...or rather a lot. I know quite a few in real life and everyone is suddenly diehard AMD and knows EXACTLY what they're talking about.

It's not everyone who supports AMD, but there just seems to be a lot more now than during the release of the A64. Maybe I just missed the posts, but the Intel camp didn't seem so upset over it.
 
I will definately buy a Conroe system, seeing as how the AM2 will only increase performance with less-than-10-%, think it was like a 3% increase or something.
 
Unless AMD gets it together in time for the AM2 release, Intel is going to steamroll them with Conroe.
 
Socket AM2, for all intents and purposes, will be pretty irrelevant at launch.
 
dr_dirtnap said:
Intel is in the position that AMD was going from the XP era to the A64 era
1. Same Memory Support
2. Very different core
3. Major architectural changes
Click to expand...
i'm gonna have to refute #2 there.. K7 and K8 are actually pretty damn similar.

conroe is really similar to dothan, which is basically a power tweaked and die shrinked p3. so really, intel is making a step backwards to go forwards here :p
 
(cf)Eclipse said:
i'm gonna have to refute #2 there.. K7 and K8 are actually pretty damn similar.

conroe is really similar to dothan, which is basically a power tweaked and die shrinked p3. so really, intel is making a step backwards to go forwards here :p
Click to expand...

I think K7 -> K8 changes are *on the surface* similar to Dothan -> Conroe. Just widening the registers and issues.

Feel free to correct me if I'm wrong.
 
(cf)Eclipse said:
i'm gonna have to refute #2 there.. K7 and K8 are actually pretty damn similar.

conroe is really similar to dothan, which is basically a power tweaked and die shrinked p3. so really, intel is making a step backwards to go forwards here :p
Click to expand...

Yeah, you are correct. I actually combined #2 and #3. The integrated memory controller, HT, and x86-64 are architectural changes not major core changes. The only major differences I recall between the actual cores are the integer and FP piplelines.

I stand corrected!
 
conroe is really similar to dothan, which is basically a power tweaked and die shrinked p3.
Click to expand...

this statement is incorrect. it is not similar to dothan. it is not a die shrink. where do you get you info?
 
(cf)Eclipse said:
intel is making a step backwards to go forwards here :p
Click to expand...

Brilliant deduction! So, AMD has always been behind Intel. All it took was for Intel to turn around and there it is. AMD flat on its ass.
 
Intel never really took a step backward but more of a 90 degree turn.
 
empoy said:
Intel never really took a step backward but more of a 90 degree turn.
Click to expand...

Oh, I agree on that one and I didn't want to add to the AMD propaganda. Fact is at the heart of all AMD and Intel processors is an instruction set that Intel created.
 
Regardless of which way they turned, it is definitely a step forward. Conroe is everything that Netburst was not:

- No longer can be used as a space heater.
- Beats AMD in the coveted performance-per-megahertz category.
- More power-efficient then ever.

:)
 
if the conroe turns out to be as good or better then its early benchmarks ill definitely get it for my next rig ,

but of course not if it will cost 400+ bucks
 
if the conroe turns out to be as good or better then its early benchmarks ill definitely get it for my next rig , but of course not if it will cost 400+ bucks
Click to expand...

Conroe offers low to high price points. bottomline, it'll be better than AM2.
 
Am I the only one who heard that AMD is Raising CPU Prices next quarter...

Do they want to lose to intel this time, or what?
 
USMC2Hard4U said:
Am I the only one who heard that AMD is Raising CPU Prices next quarter...

Do they want to lose to intel this time, or what?
Click to expand...
I heard it too. And this is while Intel is dropping prices like crazy next month.
 
USMC2Hard4U said:
Am I the only one who heard that AMD is Raising CPU Prices next quarter...

Do they want to lose to intel this time, or what?
Click to expand...

That's being discussed in the AMD forum.

Apparantly, this one is a bungled version of their infamous stick-and-carrot approach[something that is more useful for convincing reticent animals like mules]. A more successful version is the S-939 Opterons. In that ploy, they renamed the A64 and X2s "Opterons" waited about 4 months after they were supposedly "released" to actually ship them.

Some of their fan-base even made a smokescreen for them, believing the reason these chips were better overclockers was because they are meant for the server market. That doesn't make sense since CPUs aren't supposed to fail whether they're sitting in a server or a in desktop. So, the "quality for server" stuff is BS. And if people who run servers wanted to run servers with K8s and unbuffered memory they could have done it anytime prior.

In fact, AMD had enjoyed a long run with the release of the revision E chips. But, after 6 months, they experienced some stagnation, so they released their OPTERPWN or something like that. These chips were purposefully underbinned and AMD was all too happy to pay for this hype.

Now, it is a common belief that AMD knows jack about marketing. Nothing could be further from the truth. AMD markets using it's fanbase communities to create "noise".

In effect, when the commotion began to die down on Rev E and as a result of overagressive binning on the E6 chips, AMD decided to "hype it up" with the OPTERPWN or something like that. This is the carrot part of the stick-and-carrot approach. The stick part is used to beat the poor beasts. So, when the commotion became great and their inventory dropped like a rock, they jacked up the prices on these while lowering the prices on the regular A64s and X2s. In effect, the market had perceived the racket created by the fanbase communities and AMD wanted to "carrot" them in by lowering prices. But, for the fanbase, the raising of prices S939 Opteron chips acted as a good beating from a stick encouraging them to anticipate price drops, and make more noise.

Well, if you read through all that and pangs of revulsion at AMD's business practices are coursing through you, good. I'm going to give you the real explanation:

AMD will be transitioning to AM2 soon. Despite a stagnating cpu market, they will lose a little production in the short term. So, their inventory is going to dry up and they can charge higher prices. That is what they are doing.
 
Oh well, that last one took a little detour so to get back on track, I offer an opinion on the AM2 vs Conroe debate.

Conroe is a line of chips based on the NGMA. AM2 is...a socket. There's no comparison except for the comparisons between K8 and NGMA.

AMD is so shocked and awed by the results of Sring IDF that they are acting like a stunned deer caught in the headlights of a speeding Intel train. They are trying their best to spread fear, uncertainty, and doubt about Intel's new microarchitecture. But, it's not working. Reasonable fans hear only fear, uncertainty, and doubt about their new AM2 "architecture". Well, maybe just disillusionment.
 
ints great to see intel finally pushing the boundries again, however when you have competition you are always going to have a company release something that is better then its competitors so conroe may appear to be the best now but down the line AMD might release something... then later on Intel will release another new product and the cycle goes on to benifit the consumer
 
empoy said:
this statement is incorrect. it is not similar to dothan. it is not a die shrink. where do you get you info?
Click to expand...
this is what i was led to believe after reading a bit about conroe. if you can prove otherwise and correct my "incorrectness" instead of simply telling me i'm wrong, i'd appreciate it ;)
for those of you who don't know me well, there's nothing i hate more than people telling me i'm wrong without telling me why or what is actually correct.


Bao01 said:
That's being discussed in the AMD forum.

Apparantly, this one is a bungled version of their infamous stick-and-carrot approach[something that is more useful for convincing reticent animals like mules]. A more successful version is the S-939 Opterons. In that ploy, they renamed the A64 and X2s "Opterons" waited about 4 months after they were supposedly "released" to actually ship them.

Some of their fan-base even made a smokescreen for them, believing the reason these chips were better overclockers was because they are meant for the server market. That doesn't make sense since CPUs aren't supposed to fail whether they're sitting in a server or a in desktop. So, the "quality for server" stuff is BS. And if people who run servers wanted to run servers with K8s and unbuffered memory they could have done it anytime prior.

<snip>
Click to expand...
you have some interesting thoughts my friend. don't get the marketed purpose of a chip mixed up with where or how it'll actually work. most of the reasons the s939 opterons, specifically single cores, overclock so well is because this is where amd put the chips that were made in the same batches that was intended to put out FX chips, they just didn't quite make the bin for whatever reason.
for dual core opterons, it's been pure luck all along, and looking at my OC database, it appears that the X2's are actually slightly better overclockers on average :p
those people who want real servers don't use s939. end of story.
 
(cf)Eclipse said:
don't get the marketed purpose of a chip mixed up with where or how it'll actually work.
Click to expand...

Apparently, I didn't.

(cf)Eclipse said:
those people who want real servers don't use s939. end of story.
Click to expand...

Exactly!
 
Sabrewulf165 said:
I need all you fortune tellers to help me pick some winning lotto numbers.
Click to expand...
Okay, I predict that if you buy INTC, the probability of your shares doubling in value within 1.5 years is higher than your chance of winning the lotto by a factor of 2x10E8
 
Bao01 said:
Oh well, that last one took a little detour so to get back on track, I offer an opinion on the AM2 vs Conroe debate.

Conroe is a line of chips based on the NGMA. AM2 is...a socket. There's no comparison except for the comparisons between K8 and NGMA.

AMD is so shocked and awed by the results of Sring IDF that they are acting like a stunned deer caught in the headlights of a speeding Intel train. They are trying their best to spread fear, uncertainty, and doubt about Intel's new microarchitecture. But, it's not working. Reasonable fans hear only fear, uncertainty, and doubt about their new AM2 "architecture". Well, maybe just disillusionment.
Click to expand...

And you know that AMD is "shocked" because....how? I'd like you to post links/other evidence that AMD is like "a deer caught in the headlights of a speeding train" I'm an AMD fan, and I'm not shitting myself. However, I DO know that in competetive nature, they will release another product in the future that will compete on the same level or better than Intel, then Intel will do the same thing. Is no one listening? Every Intel !!!!!! here thinks that Intel just dropped the Atomic bomb of products on top of AMD, and they will never recover. If you really believe that, you don't understand the way our economy works.
 
I'm planning a complete new build in July, and, assuming AM2 and Conroe both perform similarly to how they have done in initial benchmarks, I'm either going to (obviously) build a Conroe system if the low and mid-range chips are out and available or, if they aren't out yet, get a nice, Conroe-capable motherboard and a pretty cheap Pent D to hold me over until they are out.

And yeah, Intel and AMD do appear to be switching roles. AMD plans on upping clock speeds, increasing cache sizes...the same type of thing Intel did to milk P4 until the very end. Even K8L doesn't sound too different.
 
I've may have missed this earlier but when does Conroe come out? I know AM2 comes out on 6/6/06 but when is Conroe coming out?
:confused:
 
AM2 is probably getting pushed back (based on that interview with Henri Richards), and Conroe is slated for Q3 this year. So sometime in July, August, or September (or thats the schedule).
 
SatinSpiral said:
Okay, I predict that if you buy INTC, the probability of your shares doubling in value within 1.5 years is higher than your chance of winning the lotto by a factor of 2x10E8
Click to expand...

Actually I don't play lotto, nor do I plan on buying Intel stock, but thanks for the attempt at humor.
 
live2sk8 said:
IAnd yeah, Intel and AMD do appear to be switching roles. AMD plans on upping clock speeds, increasing cache sizes...the same type of thing Intel did to milk P4 until the very end.
Click to expand...

Umm, you do realize Conroe has FOUR (4) MB of cache? Intel has hardly abandoned its strategy of increasing cache size to gain performance. And if it works, more power to them.
 
ddarko said:
Umm, you do realize Conroe has FOUR (4) MB of cache? Intel has hardly abandoned its strategy of increasing cache size to gain performance. And if it works, more power to them.
Click to expand...
They are actually "slowing down" the increase in cache. The 45nm shrink will "only" have 306 MB of L2 cache. Makes you wonder what Intel will do with the additional die space.
 
Duke3d87 said:
They are actually "slowing down" the increase in cache. The 45nm shrink will "only" have 306 MB of L2 cache. Makes you wonder what Intel will do with the additional die space.
Click to expand...

306mb of cache would be interesting. Wonder what kind of latency penalty we'd see? ;)
 
Duke3d87 said:
They are actually "slowing down" the increase in cache. The 45nm shrink will "only" have 306 MB of L2 cache. Makes you wonder what Intel will do with the additional die space.
Click to expand...
I look forward to more and more cache. :p Cache data access is faster and uses less power than external memory accesses. That doesn't matter much on desktops, but it helps on laptops.

Maybe Intel will have 16MB unified caches on 45nm quad core processors. I won't complain.
 
bobrownik said:
if the conroe turns out to be as good or better then its early benchmarks ill definitely get it for my next rig ,

but of course not if it will cost 400+ bucks
Click to expand...

Why not $400+ ? I see in your sig you are using a $500 GPU. The life of you gpu is 1/2 that of the cpu. So I don't understand what your saying. If the Conroe EE has more cache than conroe midrange I will not even think twace about guying it for $1000.

A conroe running @ 3.33 @1333 with more cache Should last a good 3 years even though Intel will be out with the much faster Peynren (spelling) by the end of 07. I will still be very happpy with my 40% slower conroe. I don't believe I will be to quik to jump on the quad core band wagon. Fact is I probably will stay with the lower cost 2 core models.

We well see how well apps. use quad core.
 
Status
Not open for further replies.
Back
Top