How will AMD answer the challenge posed by Haswell?

[Ajay]

What was the title of this thread??

 

[vshah]

I prefer the times I'm using an in-house developed tool and it spits out "Error: You should not be having this error"

I've seen one that says: "Error: if you get this, stop what you are doing and call Brad at x1500"

 

[Idontcare]

What was the title of this thread??

It is about how Intel is getting itself distracted with diversification efforts and taking its eye off of the shop, akin to what happened when they made the P4, and giving AMD another opportunity to upstage them again.

 

[Ferzerp]

It is about how Intel is getting itself distracted with diversification efforts and taking its eye off of the shop, akin to what happened when they made the P4, and giving AMD another opportunity to upstage them again.

Are they learning nothing from Cisco.....?

 

[Nemesis 1]

It is about how Intel is getting itself distracted with diversification efforts and taking its eye off of the shop, akin to what happened when they made the P4, and giving AMD another opportunity to upstage them again.

Medfield complete fail .. SB complete fail . PHi complete fail IB complete fail Haswell complete faill . Ya AMD has intell surrounded. Intel needs to just quit making transitors and go into buillding tractors

 

[Nemesis 1]

Given that he's saying he hopes for it be before Q42013, that implies that he suspects that it won't be.

It implies no such thing . It is clear Ananda is tallking about Atoms for differant usage . Intel is going after the phones first and formost. 22nm is for 2013 around 1 year after medfield release as intel shows on its roadmaps

 

[Genx87]

Guys guys you are slacking. How will AMD compete?

 

[Idontcare]

Medfield complete fail .. SB complete fail . PHi complete fail IB complete fail Haswell complete faill . Ya AMD has intell surrounded. Intel needs to just quit making transitors and go into buillding tractors

I will politely point out that you are counting Intel's chickens before they have hatched.

Medfield? Phi? Haswell?

Where is the "success"?

Those products are not success, they are attempts at success that have yet to deliver at this time.

SB was success, but that was also their mainstream venue. Failing there would have been catastrophic.

We've seen Intel attempt to get into the mobile handset market before. This is not their first attempt. I wouldn't take vegas odds on this attempt being a success until I see products (plural) sitting on the shelves at my local AT&T and Verizon stores.

XEON Phi? A good attempt at doing something with their billion dollar failure called Larrabee. Again let's not count those chickens just yet. We all know how well Intel does at projecting their imminent dominance of a market that they have yet to really enter into...

 

[pantsaregood]

How will AMD answer the challenge posed by Haswell?

 

[2is]

No... but there's also a cloud-based version. So it isn't limited and it's also not limited?



I don't think they'll pick either, frankly. Microsoft is going about it all wrong. The winRT version is cordoned off from the x86 world while the x86 version has nothing ARM-ified.

Sorry guy, but you aren't winning any debates here with your cloud argument. Before you can convince anyone to follow our lead you need to know what cloud based applications are. The fact you think that something is in "the cloud" means there are no limitations is a very clear indication you don't actually know what you're talking about.

To summarize:

ARM is NOT powerful enough to replace x86
It is NOT less fragmented
It is NOT more secure
It has far more limitations

Just about every claim you're making about ARM, the exact opposite is true.

 

[Ajay]

It is about how Intel is getting itself distracted with diversification efforts and taking its eye off of the shop, akin to what happened when they made the P4, and giving AMD another opportunity to upstage them again.

Thanks, I was starting to think it was a smart phones and tablets

Hmm, maybe AMD is onto something and that's why they hired Keller...

 

[Nemesis 1]

I will politely point out that you are counting Intel's chickens before they have hatched.

Medfield? Phi? Haswell?

Where is the "success"?

Those products are not success, they are attempts at success that have yet to deliver at this time.

SB was success, but that was also their mainstream venue. Failing there would have been catastrophic.

We've seen Intel attempt to get into the mobile handset market before. This is not their first attempt. I wouldn't take vegas odds on this attempt being a success until I see products (plural) sitting on the shelves at my local AT&T and Verizon stores.
XEON Phi? A good attempt at doing something with their billion dollar failure called Larrabee. Again let's not count those chickens just yet. We all know how well Intel does at projecting their imminent dominance of a market that they have yet to really enter into...

YA Intel had some problems with egg that was larreebbee. IF Phi catches on which it will. Thats is what makes great companies great. Take a losing investrment and turn it into billions in profit . Look at Itanic . A Intel failure that makes more money yearly than anything AMD has ever had. Medifield Is looking dam good for intels first success in phones . You guys can bet against Intel all day long . At the end of the day your all bankrupt.. In a business like Intels Success is built off of failures P4 . Now I don't call P4s unsuccessful but so many really ignorant people do.. . Itanic failure or not it gave intel a lot of knowledge of VLIW. Until we see the final curtain. What was successful or a failure in the past until the future reveals itself
Companies come and go . look at Apple top of the world at a time . Than they were over . Price bottomed smart people bought at $5. Everone else faileded. 5 dollats than gets $700 back today

 

[2is]

You guys can bet against Intel all day long . At the end of the day your all bankrupt..

Burn!

 

[NTMBK]

ARM is NOT powerful enough to replace x86

Actually, in the smartphone space where x86 and ARM are finally competing directly, ARM is so far winning.

And Sunspider is the benchmark the Xolo phone performed best in compared to its rivals.

It is NOT less fragmented

Now this bit, is very true. It's kind of ridiculous how little standardisation there is across ARM chips. Even presence of NEON isn't guaranteed, and that's seriously useful SIMD- can you imagine if an x86 chip shipped without SSE?

 

[Idontcare]

You guys can bet against Intel all day long . At the end of the day your all bankrupt.. In a business like Intels Success is built off of failures P4 . Now I don't call P4s unsuccessful but so many really ignorant people do.. . Itanic failure or not it gave intel a lot of knowledge of VLIW. Until we see the final curtain. What was successful or a failure in the past until the future reveals itself
Companies come and go . look at Apple top of the world at a time . Than they were over . Price bottomed smart people bought at $5. Everone else faileded. 5 dollats than gets $700 back today

Who is betting against Intel? The question being asked by this thread is "why would you bet on AMD".

Those two don't have to be mutually exclusive, no prerequisite for a zero-sum game here, if they do divest into appreciably different market segments.

 

[ShintaiDK]

Actually, in the smartphone space where x86 and ARM are finally competing directly, ARM is so far winning.

You mean 1 single customized ARM design leads a certain benchmark in a phone that cost 2x more?

ARM chips tends to be all over the place. Huge fragmentation and different components.

Like laptops there can be huge differences outside the CPU alone due to components.

 

[Nemesis 1]

Actually, in the smartphone space where x86 and ARM are finally competing directly, ARM is so far winning.

And Sunspider is the benchmark the Xolo phone performed best in compared to its rivals.



Now this bit, is very true. It's kind of ridiculous how little standardisation there is across ARM chips. Even presence of NEON isn't guaranteed, and that's seriously useful SIMD- can you imagine if an x86 chip shipped without SSE?

So we have another intel smartphone razr or something get those benchies . Your points are pointless ./ Intel just broke into the market give it time and it will be intels dime

 

[2is]

Actually, in the smartphone space where x86 and ARM are finally competing directly, ARM is so far winning.

And Sunspider is the benchmark the Xolo phone performed best in compared to its rivals.



Now this bit, is very true. It's kind of ridiculous how little standardisation there is across ARM chips. Even presence of NEON isn't guaranteed, and that's seriously useful SIMD- can you imagine if an x86 chip shipped without SSE?

Except I've already said on multiple occasions, I'm talking about ARM replacing x86 where x86 is the dominant force (desktops/laptops). Just like I don't expect that to happen anytime in the near future, I also don't expect x86 to replace ARM where ARM dominates anytime in the near future.

That graph actually illustrates my point pretty well. There's no magic bullet. You try and bring your power envelope down to that level, your performance goes down with it. If ARM tries to compete with i3's, i5's and i7's, it's power consumption will go up along with it and thus negate any benefit while still posing the issue of reinventing the wheel for no reason.

 

[beginner99]

Except I've already said on multiple occasions, I'm talking about ARM replacing x86 where x86 is the dominant force (desktops/laptops). Just like I don't expect that to happen anytime in the near future, I also don't expect x86 to replace ARM where ARM dominates anytime in the near future.

That graph actually illustrates my point pretty well. There's no magic bullet. You try and bring your power envelope down to that level, your performance goes down with it. If ARM tries to compete with i3's, i5's and i7's, it's power consumption will go up along with it and thus negate any benefit while still posing the issue of reinventing the wheel for no reason.

Exactly and contrary to ARM Intels current market might (or most probably) will be getting smaller and smaller over the next few decades. Hence they must find a foot in the market replacing it. It's like if a gas company would not invest in any research into solar, wind, maybe fuel cells, etc just anything energy related. I don't know if all of them do that but it is obviously a market that will at some point get smaller and smaller (100% certain) and then cease to exist completely.

Battery life on razer i looks pretty promising. But then looking at the youtube video...wow just how slow the web pages loads. would drive me nuts. I like my old school phone. Kind of funny that 72 hours of regular use seems to be good battery life.

 

[NTMBK]

You mean 1 single customized ARM design leads a certain benchmark in a phone that cost 2x more?

ARM chips tends to be all over the place. Huge fragmentation and different components.

Like laptops there can be huge differences outside the CPU alone due to components.

I'm just going to quote my other post on this topic.

It's notoriously hard to get accurate cross-platform benchmarks, and yes, Sunspider runs stock browser. But did you read my post? Sunspider was the only benchmark in the Intel phone's original review where it came out on top.

http://images.anandtech.com/graphs/graph5770/45990.png[/im]

[img]http://images.anandtech.com/graphs/graph5770/45975.png

And the single benchmark where it initially won:

And now, as you can see from my post above, that lead has been reclaimed by the iPhone 5.

If you're going to accuse me of being misleading or lying, please at least read my post properly next time.

 

[NTMBK]

So we have another intel smartphone razr or something get those benchies .

When those benches come out, I'll discuss them. But there aren't any.

The Razr I will clearly be faster than the Xolo- but this is as fast as this generation will get on the phone side. Intel has no faster clocked than 2GHz on its roadmap. They'll need to wait until Valleyview is ready to retake the lead.

Your points are pointless ./ Intel just broke into the market give it time and it will be intels dime

My points are pointless due to Intel's magic success in every new market it tries? Take a look at Larrabee and Itanium some time, or the number of years it has taken them to get into smartphones (Moorestown was a total flop).

 

[Arzachel]

*pics*

Thanks for posting the benches, Krait seems to stomp all over nearly everything. Might have to get a Nexus G/LG Optimus G if the price is right.

 

[ShintaiDK]

I'm just going to quote my other post on this topic.

And you should read the comment in the other thread too.

 

[NTMBK]

And you should read the comment in the other thread too.

Hah, yeah, I feel a little bad for the amount of duplication going on. Also the amount of threadjacking... what was this topic called again? :whiste:

 

[Nemesis 1]

Well the title is how does AMD compete with Haswell .This is old thinking. AMD is not intels comp. any longer its arm . So AMD just doesn't matter period . NV is a greater threat to intel than AMD . The Part no one is looking at is in smart phones INTEL is not a Monopoloy company.. IN this space I am of the opion that Intel needs an intel branded smart phone .. and as in the past .. Intel needs to work on their intel inside promos. Helping other companies paying for advertizing on tV or where ever. As it is right now no one cares whats inside of a smartphone . But with NV in the game that will change and intel will also try to sale intel inside. With windows 8 I am not so sure intel is a monopoly any longer . In the tablet space and the Smartphone space intel is not a monopoly. This allows intel to change its business model for this space . Intel can undercut everyone so long as they show a profit. Than when they are a monoploly again .. they can raise margins.