Is AMD gonna have an answer to Conroe.

[sandorski]

Originally posted by: Henny

Originally posted by: sandorski

Originally posted by: Avalon
Not immediately. What you see from AMD now is what they'll have when Conroe starts hitting the market.

Although, they do have 65nm CPUs due out late Q4 of this year, which may be interesting. If 65nm does anywhere near for AMD what it did for Intel, it will definitely make their chips more competitive.

AMD is also going to go through a string of price drops on their A64 and X2 line of processors over the next coming months, if you can call that an "answer".

Re the Price Cuts: I think once AMD cuts prices the Price/Performane differences between AM@ and Conroe will likely keep AMD competitive for awhile. AMD would certainly want to be the undisputed Performance Champ, but being the Price/Performance Champ isn't too bad, especially for a smaller Player in the market(assuming of course that Profits can be maintained at rates high enough to carry out R&D and other necessary expenses).

That's the issue. Price drops from AMD will be very painful since they they don't even have the benefits of 65 nm technology yet.

The majority of Intel's CPU's will be on 65 nm before AMD ships it's first 65 nm part. That has huge P&L implications not to mention Intel's economies of scale.

AMD can neutralize the price/performance advantage of Intel's new offering but it'll likely send them deep into the red and they need about $2.5B to convert their 200mm Dresdon fab to 300mm

I think we'll see AMD sacrifice destop market share but do all they can in the high end server space where they'll still have an advantage over Intel.

AMD has pretty much stated that's what they'll be doing(last point you made). AMD has been in this position before, so if anyone can survive under these circumstances it would be AMD. They'll come back with something eventually, with their growth in the Server/Workstation Market they may even not lose Profits, but actually make more. Time will tell.

 

[Viditor]

Originally posted by: Henny



That's the issue. Price drops from AMD will be very painful since they they don't even have the benefits of 65 nm technology yet.

The majority of Intel's CPU's will be on 65 nm before AMD ships it's first 65 nm part. That has huge P&L implications not to mention Intel's economies of scale.

1. Can you explain your viewpoint as to what those implications are? 65nm in and of itself does not make a processor less expensive to manufacture...what makes it cheaper is if the dies size is smaller. The Manchester cores are 147 mm2, and the Presler is 162 mm2...so the Presler (at 65nm) is actually more expensive in terms of die size than the X2 Manchesters at 90nm.

2. Since we still don't know the process of the new generation Intel CPU (or even the die size), nobody can tell if it will be less or more expensive. However since it's the beginning of the ramp it is sure to be more expensive for several months.

3. In a transition like this, the economies of scale actually work against Intel. This is because they must make much larger investments just to maintain the status quo.

AMD can neutralize the price/performance advantage of Intel's new offering but it'll likely send them deep into the red and they need about $2.5B to convert their 200mm Dresdon fab to 300mm

I think we'll see AMD sacrifice destop market share but do all they can in the high end server space where they'll still have an advantage over Intel.

1. Even at the new prices, AMD is making a much better profit than Intel. The depreciation of the equipment acts as surplus cash because they've delayed 65nm production. In other words, the longer they can use 90nm equipment and still remain competitive, the cheaper it is to manufacture CPUs for them.

2. AMD had $2.63 Billion in cash last quarter, and most assuredly has more now. They also have East Germany financing most of the new Fab at very low interest.

3. There is no need for AMD to sacrifice anything to maintain and even increase their marketshare in servers...but I don't understand why you think that cutting prices on desktop parts will cause AMD to "sacrifice desktop market share". It seems to me that they are doing exactly the opposite of this!

 

[sandorski]

Originally posted by: Viditor

Originally posted by: Henny



That's the issue. Price drops from AMD will be very painful since they they don't even have the benefits of 65 nm technology yet.

The majority of Intel's CPU's will be on 65 nm before AMD ships it's first 65 nm part. That has huge P&L implications not to mention Intel's economies of scale.

1. Can you explain your viewpoint as to what those implications are? 65nm in and of itself does not make a processor less expensive to manufacture...what makes it cheaper is if the dies size is smaller. The Manchester cores are 147 mm2, and the Presler is 162 mm2...so the Presler (at 65nm) is actually more expensive in terms of die size than the X2 Manchesters at 90nm.

2. Since we still don't know the process of the new generation Intel CPU (or even the die size), nobody can tell if it will be less or more expensive. However since it's the beginning of the ramp it is sure to be more expensive for several months.

3. In a transition like this, the economies of scale actually work against Intel. This is because they must make much larger investments just to maintain the status quo.

AMD can neutralize the price/performance advantage of Intel's new offering but it'll likely send them deep into the red and they need about $2.5B to convert their 200mm Dresdon fab to 300mm

I think we'll see AMD sacrifice destop market share but do all they can in the high end server space where they'll still have an advantage over Intel.

1. Even at the new prices, AMD is making a much better profit than Intel. The depreciation of the equipment acts as surplus cash because they've delayed 65nm production. In other words, the longer they can use 90nm equipment and still remain competitive, the cheaper it is to manufacture CPUs for them.

2. AMD had $2.63 Billion in cash last quarter, and most assuredly has more now. They also have East Germany financing most of the new Fab at very low interest.

3. There is no need for AMD to sacrifice anything to maintain and even increase their marketshare in servers...but I don't understand why you think that cutting prices on desktop parts will cause AMD to "sacrifice desktop market share". It seems to me that they are doing exactly the opposite of this!

"East Germany", whoa AMD has gone back in time!

Ya, ok, what you said is probably true. It just struck me as funny.

 

[Praxis1452]

amd 65nm rev G b4 the end od the year

 

[DidlySquat]

Originally posted by: Noubourne
K8L 2007. Have patience. They've been winning for 2 years now. They can take 9 months off.

no, the conroe release with no answer from AMD for AT LEAST 6 MONTHS (and even then it will only be a small speed bump for cherry picked high end FX chips) while Conroe has enourmous OC potential, really marks a turn of the tide, with Intel probably looking being the top performer for quite a few years to come. AMD is just going from bad to worse, and even K8L will probably not help them turn the battle. Maybe K9 but again the Conroe architecture is such a leap i think it will be at least 3 years before AMD might be able to catch up.

 

[TuxDave]

Originally posted by: zephyrprime

Originally posted by: TuxDave

Originally posted by: OcHungry

Originally posted by: inveterate
Haven't heard anything.. I'm scared of having to buy all new stuff.

AM2 is the way to go. K8l and 65nm's are socket AM2 and it will only get better.
The Conroe hype is cooling off as time goes by and independant reviews conducted.
This website has conducted a few multitask benchmarks and the results are not so great.
The FSB Bottlenecks in multytask and dual GPU.

I'm kinda wondering if his 'multitask' benchmark becomes a bigger limitation on HDD access. What is the 'other' benchmark he's running and where is the A64 benchmark for comparison?

He's also running the benchmark on a P4 XE and the conroe doesn't look so good compared to it even though the XE would naturally be affected by HD performance as well.

Which is why if the bottleneck is not the processor, then the scores would end up to be similiar like what we're seeing on the website.

 

[Accord99]

Originally posted by: OcHungry
This website has conducted a few multitask benchmarks and the results are not so great.
The FSB Bottlenecks in multytask and dual GPU.

It's up against the P4XE, which can handle up to 4 threads simultaneously and gain a throughput increase. Versus a A64 X2, Conroe multi-tasks better.

http://www.xtremesystems.org/forums/showpost.php?p=1419278&postcount=1