Jim Keller joins Samsung

[NTMBK]

http://www.overclock3d.net/articles/cpu_mainboard/amd_s_kim_keller_reportedly_joins_samsung/1

 

[geoxile]

RIPIP everyone else

If this is true anyway. I wonder just how much funding Samsung would be willing to give Keller and co. Perhaps they know something about K12?

 

[Arachnotronic]

RIPIP everyone else

If this is true anyway. I wonder just how much funding Samsung would be willing to give Keller and co. Perhaps they know something about K12?

LOL, do people really think that Jim Keller is "chip god"?

 

[ShintaiDK]

LOL, do people really think that Jim Keller is "chip god"?

Only hype believers.

Joining Samsung already explains the fast exodus from AMD.

 

[Arachnotronic]

Only hype believers.

Joining Samsung already explains the fast exodus from AMD.

Agreed.

 

[Skurge]

RIPIP everyone else

If this is true anyway. I wonder just how much funding Samsung would be willing to give Keller and co. Perhaps they know something about K12?

A substantial amount I would imagine if you have the talent, fabs and money. You could beat anyone. Especially the way Intel is sandbagging in mobile.

 

[dark zero]

A substantial amount I would imagine if you have the talent, fabs and money. You could beat anyone. Especially the way Intel is sandbagging in mobile.

RIP Intel on Mobile then.... Mediatek is cheap, Samsung is the MT king and Apple is the ST king.... Intel has nothing to do here.

 

[geoxile]

A substantial amount I would imagine if you have the talent, fabs and money. You could beat anyone. Especially the way Intel is sandbagging in mobile.

From the leaks regarding Mongoose Samsung is already on track to produce the new king of mobile SoCs.

 

[NTMBK]

Given the success of Apple's A series, it makes sense. Have they picked up anyone else from that team?

 

[Arachnotronic]

From the leaks regarding Mongoose Samsung is already on track to produce the new king of mobile SoCs.

There are no leaks of Mongoose. There are just some random posts on Weibo from sources that have proven completely unreliable.

 

[Boze]

Only hype believers??

You youngsters need to stop in and talk to people who've been at Intel for 10+ years.
Stop in and talk to people who worked with him at DEC.
Stop in and talk to people who worked with him at AMD.

He understands modern central processing unit design unlike anyone on the planet.

If you don't believe that, you need only look at where Apple was with their A-series chips before Jim came onboard to work his magic.

 

[Arachnotronic]

Only hype believers??

You youngsters need to stop in and talk to people who've been at Intel for 10+ years.
Stop in and talk to people who worked with him at DEC.
Stop in and talk to people who worked with him at AMD.

He understands modern central processing unit design unlike anyone on the planet.

If you don't believe that, you need only look at where Apple was with their A-series chips before Jim came onboard to work his magic.

Have you actually talked with such people? I have no doubt that Keller is a good engineering manager, and before he was a manager was a good CPU architect, but the hyperbolic claims that he's a "legendary" CPU architect and that everything he touches turns to gold get really tiresome really quickly.

The A-series chips are the product of many very smart engineers working with a very large budget in service of a very specific, highly focused product.

 

[ShintaiDK]

The A series is a good example. Jim Keller doesn't work there anymore, and they still make even better products. In other words, Jim Keller was just 1 random person out of 1000s. Is he better than the average? Sure, without question. But Jim Keller isn't even replacing 10 other "unknown" engineers and managers in the grand scheme.

 

[Fjodor2001]

The A series is a good example. Jim Keller doesn't work there anymore, and they still make even better products.

There's a long lead time between when an uArch design is done until it shows up in actual products on the market. We're talking several years. So the work he did at Apple some years ago may still have an impact on the products we're seeing today.

Also, uArch design principles and practices established at the time he was at Apple can still have impact on the work they have been doing on the follow-up designs to the ones he was directly involved in.

 

[ShintaiDK]

There's a long lead time between when an uArch design is done until it shows up in actual products on the market. We're talking several years. So the work he did at Apple some years ago may still have an impact on the products we're seeing today.

Also, uArch design principles and practices established at the time he was at Apple can still have impact on the work they have been doing on the follow-up designs to the ones he was directly involved in.

Its over 3 years since Keller left Apple. The assumption that he is a god is hilarious at best.

 

[Fjodor2001]

Its over 3 years since Keller left Apple.

Yes, I know. So the work he did can still have an impact on the current Apple CPUs, as I wrote in my previous post.

 

[simboss]

Also, uArch design principles and practices established at the time he was at Apple can still have impact on the work they have been doing on the follow-up designs to the ones he was directly involved in.

:thumbsup:

If you want a public example of what a good base and design principles from the start can give you, you can look at open-source SW projects as well.
The most successful projects will always involve 100s or 1000s of people in the end, but they have been created and put into the right direction by a few, and sometimes even by one.

Also having the right person at the top often means he can identify and use the right persons below him and that organization will stay in place unless someone of something disrupts it.

 

[Tuna-Fish]

I don't personally think that Keller is a chip-design Midas, rather than just a really competent designer, but...

Its over 3 years since Keller left Apple. The assumption that he is a god is hilarious at best.

The swift core took more than 5 years from start of design to silicon on the market. Assuming the high-level team has kept up with the releases, A9 high level design was just about finished when Keller left.

The rapid pace of releases makes people forget just how long it takes to design a microprocessor from start to finish. Everyone who tries to keep up with the state of the art needs to have multiple projects in different state of completion going on simultaneously. This is why if you find that your chip is going in the wrong direction (P4, BD), it takes several releases before you can correct your mistakes. Prescott designers did not know the real troubles that Northwood faced before the design was finished. These days, Intel has two completely separate chip design teams that work on alternate uarchs, and both teams also always have two projects going on, so they are building 4 uarchs at the same time, just for their primary CPU line.

This is also why Zen+ keeps being mentioned in ex-AMD bios. If they want to release and update to Zen a year or two after release, high-level design on it needs to already be done.

 

[stuff_me_good]

The A series is a good example. Jim Keller doesn't work there anymore, and they still make even better products. In other words, Jim Keller was just 1 random person out of 1000s. Is he better than the average? Sure, without question. But Jim Keller isn't even replacing 10 other "unknown" engineers and managers in the grand scheme.

Pretty simple minded thinking if you ask me. I'm not Keller expert, nor do I know if he is chip god or not and how much better he is than the average engineer, but have you ever considered that if he is as good as some here are claiming, he put the long lasting seed to grow in fruition while working on apple?

Maybe he draw future designs in apple and lead them to design right kind product far into the future while he was gone and not to make faildosers or stupid thing like that?

 

[PPB]

At this pace, Intel will only be competing with Keller's designs made at different companies. Props to him and his ability to influence his field of expertise.

Obviously the mediocre will only bash him, at Apple he left exactly 1 year before the launch of A7 producs. Quite telling that he left that company with that design under his belt. Considering A8 and A9 are evolutionary steps and not radical redesigns, this is quite telling of his metedology (lay a good fundation with still low hanging fruit to make quick, evolutionary follow ups).

To say he is 1 random person out of 1000, my god, how someone can be this asinine?

 

[krumme]

Put him on the ignore list. Constantly he gets the threads derailed.

 

[Madpacket]

Who's a better chip designer than Keller?

Suddenly I hear the sound of crickets...

 

[ShintaiDK]

Who's a better chip designer than Keller?

Suddenly I hear the sound of crickets...

How did you ever hear about Keller compared to all the ones we dont hear about?

Nobody here cant really tell why Keller should be better. Besides he was one of the people working on K7, K8, A4 and A5.

 

[Arachnotronic]

How did you ever hear about Keller compared to all the ones we dont hear about?

Nobody here cant really tell why Keller should be better. Besides he was one of the people working on K7, K8, A4 and A5.

I bet you that not a single person here waxing lyrical about the greatness that Jim Keller is even knew the man existed before AMD put out a PR touting that it had hired him back in 2012.

I wonder if these people know the names of the folks who have led various successful projects across the industry? There are a lot of smart people in this industry and a lot of really great chips get made. I guess you need to get hired by AMD in a bid to revive a dying CPU operation before you get any public recognition I suppose...

EDIT: to be clear, I'm not trying to "put down" Keller -- his career history points to somebody who is extremely good at what he does. I'm just trying to say that the people saying "OMG RIP ANYBODY COMPETING WITH KELLER" grossly underestimate the talent pool within the very large and vibrant semiconductor industry.

 

[simboss]

I bet you that not a single person here waxing lyrical about the greatness that Jim Keller is even knew the man existed before AMD put out a PR touting that it had hired him back in 2012.

I wonder if these people know the names of the folks who have led various successful projects across the industry? There are a lot of smart people in this industry and a lot of really great chips get made. I guess you need to get hired by AMD in a bid to revive a dying CPU operation before you get any public recognition I suppose...

EDIT: to be clear, I'm not trying to "put down" Keller -- his career history points to somebody who is extremely good at what he does. I'm just trying to say that the people saying "OMG RIP ANYBODY COMPETING WITH KELLER" grossly underestimate the talent pool within the very large and vibrant semiconductor industry.

Very true as well, a very talented engineer can do a massive and long lasting difference, but mostly if he can hire and influence a good team around him, which takes time and money, I am not sure he had any of these at AMD, there is a better chance he gets them at Samsung though.