Conroe vs. X2 predictions

[Hulk]

Okay, I should have made this more clear. Here are the "rules":

1. Predict what will be the fastest desktop offerings from Intel and AMD when Conroe launches.

2. Predict their performance in games, office applications, and audio/video/graphics applications.

Everybody in here has lots of knowledge and opinions. I'm just saying let's put them in writing. It'll be fun to have a look back at this thread upon launch.

My predictions.

The chips will be Conroe at 2.67 and AMD FX62 at 2.8GHz

Conroe 2.67 vs. FX62

Games - FX62 by a narrow margin

Office Applications - FX62 wins bya narrow margin

Audio/Video/Graphics/Multimedia - Conroe will by a narrow margin

 

[Furen]

I think you'd be better off comparing the 2.6GHz X2 5000+ to the Conroe 2.67, which should have a similar price, too. Also, dont forget that there will be a Conroe 3.33GHz Extreme Edition, which I think will be the (overall) best CPU out there (not sure if it'll actually be released at the same time, though).

 

[Hulk]

I'm comparing what we've read will actually be released by Q3. I haven't seen Conroe EE at 3.33GHz on any official Intel roadmaps.

If you have please post a link I'd love to see it!

 

[robertk2012]

The new Conroes will be tuff to beat and I think Intel will take the lead then. For how long I dont know. AMD better have something up their sleave. The new Core Duos at 2 ghz are a strong addition. I cant really imagine a processor with that architecture running at 3.33. Fun times we are in.

 

[Hulk]

And I forgot to add, that is part of the prediction. Actually predicting what chips will be available at Conroe launch.

I think Conroe at 2.67 will be top of the desktop line for Intel and FX62 top of the desktop line for AMD.

I should have noted that.

 

[Furen]

The Conroe 2.67 is going to be like 630 bucks, though, so comparing it to the fx-62 (even if there's no XE at launch) is kind of unfair.

robertk2012: Yonah at 2GHz is slightly slower than an X2 with 512k L2/core at the same clock so if Conroe only matches Yonah in performance per clock it'll still be behind AMD's 2.6GHz 1mb/core chips. Conroe is going to be significantly different from Yonah (I think) so I'd expect it to perform a bit better per clock. If Conroe ends up basically being a 64bit Yonah then AMD will be in damn good shape for a while yet.

 

[robertk2012]

Conroe is expected to remove certain power constraints and probably widen thermal envelope. It wont be the same architecture but will definately not be netburst. Just think of how much less power a core duo uses than an A64 to get the same performance. Im thinking that Conroe will be a beast.

 

[dexvx]

Why would Conroe be better at multimedia/encoding, etc?

The reason the Pressler is so good is because it has a high clock and can do more SSE instructions. I doubt the Conroe would be as good in those than even an overclocked Pressler

 

[robertk2012]

The intel architecture has shown itself to be better at multimedia and encoding. Thats one of the reasons apple went with intel. AMD will probably maintain the lead in gaming though. Its probably because of the intergrated memory controller.

 

[Hulk]

Come on guys, just make the predictions!

I think Conroe will be good at multimedia for a few reaons:

Rumor has it Conroe will have an improved floating point unit. Intel has a good track record of making good fpu's when they want to. And I think they want to now.

SSE4 - I have a hunch it and other optimizations will make a difference for Conroe right off the bat.

The 14 stage pipe will stall less with multimedia than with integer, so AMD with a 12 stage pipe won't have the efficiency advantage it might have.

As I said it's just one person's educated guess over another's at this point.

Put your cards on the table!

 

[dmens]

My predictions: on average, 90/100 perf ratio on same frequency, X2 vs merom. merom will be better in float/SSE, X2 will be better with ints, probably negligible (but not sure on that). On the same performance, merom will be draw about 75-80% of the power of the X2.

 

[Marmion]

Originally posted by: dmens
My predictions: on average, 90/100 perf ratio on same frequency, X2 vs merom. merom will be better in float/SSE, X2 will be better with ints, probably negligible (but not sure on that). On the same performance, merom will be draw about 75-80% of the power of the X2.

So does that means conroe on a faster bus will beat X2 easily?

 

[Cooler]

I think the uber high end 2.93 and 3.33 ghz models will come out on top while low mid range will be even with what amd has to offer.

 

[openwheelformula1]

knowing that current Yonah almost matches AMD clock for clock, I would say Conroe will have an advantage. Also note that AMD's rev.F will improve performance and efficiency slightly. The key is Conroe, will it dissapoint or will it "leap forward"? There are goods such as SSE4 that interests me, but there are bads such as lack of current chipset support. One thing that AMD has really done well is the path of upgradability.

 

[dmens]

Originally posted by: Marmion
So does that means conroe on a faster bus will beat X2 easily?

imo that is correct.

 

[n7]

I think clock for clock, Conroe will be very very close to the X2s.

I actually think the Conroe will be better for games, & the X2s better with media/video encoding, etc., since with Yonah, that was generally how it seems to stack up.

It does seem like Intel is going to be able to ramp up speeds faster than AMD, & ironically enough, it appears this will be a clockspeed battle.

 

[dmens]

the X2s better with media/video encoding, etc.,

yonah was deliberately crippled on FP/SIMD. merom rectified that situation.

 

[Quinton McLeod]

I predict the Conroe still won't win over AMD. I automatically declare Conroe the best is just as bad as the PSP vs. DS threads in video game forums. In the past, people automatically predicted that PSP will be the best and overthrow Nintendo. Nowadays, it's the exact reverse.

With Intel and AMD, the Conroe will not beat AMD by much. If anything, they'll be tied.

 

[openwheelformula1]

one thing is for sure though, AMD will still be the better bang for buck due to bias public perception.

 

[dmens]

I predict the Conroe still won't win over AMD... the Conroe will not beat AMD by much. If anything, they'll be tied.

<, > and == in one post, LOL.

P.S. I'm not guessing. I'm sticking with my landing zone.

 

[Hulk]

Come on guys. This is simple. Make a guess and stick with it.

Okay, I should have made this more clear. Here are the "rules":

1. Predict what will be the fastest desktop offerings from Intel and AMD when Conroe launches.

2. Predict their performance in games, office applications, and audio/video/graphics applications.

 

[BrownTown]

Conroe's strength will most likely be integer operation, and it will be relatively weak at floating point operations (although not as crappy as Yonah). Likely Conroes strengths will be games and synthetic type benchmarks, but will do worse to encoding/multimedia. SSE4 means nothing to anyone since you have to code to exploit it, and since AMD doesnt have it yet no benchmarks or games will use it. It will be years before SSE4 has any impact on performance

 

[robertk2012]

1. Im going with the Intel pentium 166
2. They will all suck with Vista

 

[Hard Ball]

Originally posted by: dmens
My predictions: on average, 90/100 perf ratio on same frequency, X2 vs merom. merom will be better in float/SSE, X2 will be better with ints, probably negligible (but not sure on that). On the same performance, merom will be draw about 75-80% of the power of the X2.

Your prediction is almost certainly wrong, it is likely to be the other way around; Merom will excel in ALU, while X2 at FP; as in regard to SIMD units, it's way early to tell, since we don't really know the full specs of SSE4 yet, when each company is likely to implement these.

 

[dmens]

Your prediction is almost certainly wrong, it is likely to be the other way around; Merom will excel in ALU, while X2 at FP

Care to explain? By the way, I'm talking about real large workloads, not dinky little ones that fit in the cache.