Dr. Lisa Su accounced as AMD’s new president and CEO

[Monkeynut]

Let's hope RAMBUS doesn't come back out of nowhere and try to Su.

 

[R3X]

Not nearly hot enough

 

[provost]

compared to what?


lol

 

[Final8y]

Shes his sister.

 

[R3X]

Both the same person he just grew his hair out and got re-branded to a dr, even back then he was struggling to hold the AMD gpu.

 

[Wizzfizz]

They are the same person...

 

[Orangey]

Jensen got dem mussles doe

 

[Ayahuasca]

and he's a bloke, men in charge > women in charge.

Sorry feminists.

 

[humbug]

Some spinets from http://seekingalpha.com/article/2557375-thoughts-on-amds-new-ceo?uprof=55

In some aspects, such as technology, it feels as if Rory is limited in what he can do as it is fairly clear that he has little knowledge on the technical aspects of the very complicated business. Through various interviews, Rory has given us nearly nothing but repetitive information, and Wall Street has been fed up with it. Technical questions have either been handed off to more capable people like Lisa, or entirely avoided.

Rory was never an engineer, a good strategist and leader yes, but i guess you need to be all things at that level, i think Lisa fits those boots.

In a recent interview, Lisa stated that she recognizes her team's talents and how important it is to empower them to innovate and get things done "simply, fast and decisively." Releasing products on time is going to become crucial to AMD's survival as it is running out of oxygen. AMD has consistently suffered from past delays on CPUs and APUs, and the company may not have the time to continue down that road.
AMD will go big or go home in 2016.

Right, exactly! stop diverging and get on with it.

2016 will be an important year for AMD, as we will see the successors to the failed Bulldozer architecture that got Rory Read's predecessor fired in 2011. Dubbed, K12 and Zen, these new designs will make or break AMD. Rory in a recent interview:

"Everyone knows that Bulldozer was not the game changing part when it was introduced three years ago. We have to live with that for four years but Zen, K12 we went out and got Jim Keller, we went out and got Raja Koduri from Apple, Mark Papermaster, Lisa Su. We are building now our next-generation graphics and compute technology that customers are very interested in and they'll (referring to the next generation graphics and compute architecture) move to the next generation node and they'll be ready to go."

So the man responsible for Bulldozer got fired? i didn't know that.
I think its right to say K12 will be make or break for AMD, and i think its good to know AMD themselves think of it like that, it will give them focus, they will throw everything at it and we might just end up with another AthlonXP moment, AMD are capable of it, especially with Jim Keller leading the charge again.

 

[mmj_uk]

Any new CEO coming in will impress everybody with talk, the question will ultimately be can she deliver with the resources at her disposal?

I remember there was an ex-AMD employee talking on the Apple forums before Bulldozer was released and he said how rubbish it would be when everyone still thought it would be amazing (and he got attacked on this forum), he explained how AMD had got shut of a lot of their best engineers and had started automating a lot of their work which was much less efficient (in terms of die organisation/space) than doing things by hand.

It sounds like AMD have been in decline behind the scenes for a very long time so she is going to have to do amazingly well for AMD to have another K7 on their hands come 2016, it's not impossible of course but AMD have fallen so behind Intel it's not even funny, if AMD even made a CPU as powerful as 5820K today it would probably have a TDP of 300W.

 

[LambChop]

To add, I think they will have to throw a hell of a lot of money at it too.

 

[CAT-THE-FIFTH]

AMD is making two new cores - Zen and K12 and AFAIK they will share R and D between them. Basically they are cutting down from 4 supported lines to 2. Carrizo will be the last of the current generation of FX CPUs with the BD derived cores(Excavator).

Personally I see AMD aiming more towards the sub £150 area in consumer X86 CPUs,ie,things like APUs with improved performance/watt and performance/MM2 than the current lot,and going more towards designs they can plonk into laptops and lower power desktops. APUs already make up around 75% of total AMD CPU sales,and it is rising each year.

I doubt we will see true successors to the Phenom II X6 and FX83** line ever again - I would be surprised if they did make one.

The K12 cores looks to be more towards tablet and microserver use it appears.

 

[Orangey]

Say hello to £1000 2600K-level parts if so.

 

[humbug]

AMD is making two new cores - Zen and K12 and AFAIK they will share R and D between them. Basically they are cutting down from 4 supported lines to 2. Carrizo will be the last of the current generation of FX CPUs with the BD derived cores(Excavator).

Personally I see AMD aiming more towards the sub £150 area in consumer X86 CPUs,ie,things like APUs with improved performance/watt and performance/MM2 than the current lot,and going more towards designs they can plonk into laptops and lower power desktops. APUs already make up around 75% of total AMD CPU sales,and it is rising each year.

I doubt we will see true successors to the Phenom II X6 and FX83** line ever again - I would be surprised if they did make one.

The K12 cores looks to be more towards tablet and microserver use it appears.

The reason APU's are 75% of their market is because they haven't renewed their CPU's in 3 years.

If AMD made a more efficient higher performing 6/8 Core CPU they would sell.

 

[CAT-THE-FIFTH]

Say hello to £1000 2600K-level parts if so.

The thing is the Brazos platform ended up being an unexpected success for AMD. It was a low cost small core,designed with a limited budget made on an old process and AMD sold over 30 million of them. It had extensive use of automated tools too.

Fast forward to Jaguar and they have sold millions of them in consoles and low cost computers.

Again another small core which was quite efficient in R and D and production costs.

http://www.anandtech.com/show/7990/amd-announces-k12-core-custom-64bit-arm-design-in-2016

Jim Keller added some details on K12. He referenced AMD's knowledge of doing high frequency designs as well as "extending the range" that ARM is in. Keller also mentioned he told his team to take the best of the big and little cores that AMD presently makes in putting together this design.

So what I see from what has been said so far,is that Zen will be an intermediate size modular core,which can run at higher frequencies(probably the 3GHZ to 4GHZ),while being quite efficient in performance/watt and probably performance/mm2. It will probably share parts with the K12 ARM core.

Whether it will exceed Intel with their "large cores" is another thing,but remember AMD is at the mercy of its fab partners and Intel tends to be a node ahead/transitions to more advanced nodes quicker than AMD. So they need to be able to tackle that.

Its less performance more performance/watt and performance/mm2 which is hurting AMD for CPUs like the FX8350 or even their Kaveri chips. Intel is simply selling smaller chips with higher profit margins than AMD(even with process node costs included),and they can get higher performing chips into more compact desktops and laptops too.

The reason APU's are 75% of their market is because they haven't renewed their CPU's in 3 years.

If AMD made a more efficient higher performing 6/8 Core CPU they would sell.

The traditional server market was what was sustaining the AM3+ socket and before,but they appear to be retreating from that now,hence why they have not updated the line.

However,if you look at the AMD strategy,with the purchase of SeaMicro,its more about using X86/ARM low power cores and using many of them.

These are the kind of platforms AMD is now targeting:

http://www.amd.com/en-us/press-releases/Pages/arm-cortex-a57-based-2014sep30.aspx

Its not to say that a Zen based quad core would be slow but I am not sure it will be an Intel beater though in pure performance though,and ultimately they don't really need to IMHO. They need to improve margins and improve performance/watt.

Remember enthusiasts worry about 15% to 20% extra performance here and there,but for most of the market its not really as important as people think.

Intel probably sells far more Pentiums and Core i3s than Core i5s I suspect worldwide each year.