AMD CPU speculation... and expert conjecture

[JAYDEEJOHN]

Not for nothing I'm still waiting for all the profits AMD is going to make from GF
and all the new customers, While Intel go broke from being fab heavy.

Since you feel inclined on old news
http://www.tomshardware.com/news/amd-mubadala,19550.html
As I said long ago.
As for Intel heavy in fab, this new slower market has taken them by surprise, and changes have been made already.
Now, as far as sinking ships, look to pirate bay for that, as only ThunderMan had those visions

 

[noob2222]

amd says they will keep supporting socketed cpus
http://techreport.com/news/24009/amd-were-not-abandoning-socketed-cpus
tbf, intel never said they will abandon socketed cpus entirely. chances are hedt platform will retain lga cpus (thus forcing interested users pay moar moniez).

Intel is thinning the motherboard vendors. they are abusing their market position to force companies out of business. Isn't Intel such a great company to have?

companies like ECS, Evga, biostar, foxconn, and jetway aren't going to be able to afford buying up hoards of Intel processors to solder onto their product and wait till it sells.

Its questionable to see who comes out on top, asus, msi, gigabyte, asrock ... Think about OVERSTOCK. not only would they lose money on the boards that didn't sell, now they are giving intel money for a cpu that didn't sell.

Intel is trying to turn motherboard vendors into pc vendors such as dell, hp, and acer.

Its a great move for Intel's pocketbook, they get to sell one cpu for every motherboard manufactured. Its a disaster waiting to happen for small business computer & repair shops.

How many businesses will go away at the hands of Intel's 14nm innovation?

 

[earl45]

Since you feel inclined on old news
http://www.tomshardware.com/news/amd-mubadala,19550.html
As I said long ago.
As for Intel heavy in fab, this new slower market has taken them by surprise, and changes have been made already.
Now, as far as sinking ships, look to pirate bay for that, as only ThunderMan had those visions

You have a history of making very biased hypothesis, for AMD and against Intel,
which I may add has not taking place at all to date.
So after 3 years or more of doom and gloom for Intel and AMD being so smart about GF
you have been wrong about it all.

 

[Cazalan]

Which server chips are you referring to?

ARM blades mostly.

 

[Cazalan]

How many businesses will go away at the hands of Intel's 14nm innovation?

If the socket was key to their business then they're on the decline already.

They're ignoring the tablet/phone/netbook and other embedded markets.

 

[twelve25]

ARM blades mostly.

http://regmedia.co.uk/2011/10/30/calxeda_energycore_server_board.jpg

I figured it had to be something like that. I was scratching my head trying to remember which new Opteron or Xeon slot processor I had missed.

 

[de5_Roy]

Globalfoundries to Start 14nm Tests in Q1 2013 - Company.
AMD Will Not Benefit from Globalfoundries Rapid Transition to 14nm
http://www.xbitlabs.com/news/other/display/20121204221042_Globalfoundries_to_Start_14nm_Tests_in_Q1_2013_Company.html

 

[JAYDEEJOHN]

You have a history of making very biased hypothesis, for AMD and against Intel,
which I may add has not taking place at all to date.
So after 3 years or more of doom and gloom for Intel and AMD being so smart about GF
you have been wrong about it all.

I said they had a partner, so my link.
I also said Intels investments in their fabs will costs them, as they ratchet down.
Their current stock position is not what it was, and is my point.

 

[noob2222]

If the socket was key to their business then they're on the decline already.

They're ignoring the tablet/phone/netbook and other embedded markets.

as I stated, Intel is trying to force motherboard vendors into the likes of hp, dell, acer, ect.

motherboards are cheap to make in comparison to a full computer.

small computer repair shops will be hit the hardest ... "to fix your dead cpu, this will cost $500 for the motherboard with the soldered cpu" ... you can buy a new computer for that, customer leaves and goes to best buy.

 

[earl45]

I said they had a partner, so my link.
I also said Intels investments in their fabs will costs them, as they ratchet down.
Their current stock position is not what it was, and is my point.

Let's see AMD sold it's fabs to GF and you said that was a good thing being fabless
you turn out to be completely wrong on this.
This is what GF gave AMD.
1) All CPU's delayed to market and under performing.
2) Yield problems.
3) The cost of doing business with GF has cost AMD dearly.

Let's not forget GF being ahead of Intel in process,
you were more wrong then ever right.

 

[noob2222]

Let's see AMD sold it's fabs to GF and you said that was a good thing being fabless
you turn out to be completely wrong on this.
This is what GF gave AMD.
1) All CPU's delayed to market and under performing.
2) Yield problems.
3) The cost of doing business with GF has cost AMD dearly.

Let's not forget GF being ahead of Intel in process,
you were more wrong then ever right.

how much money does AMD have to upgrade the fabs?

How much money does Abu Dhabi (aka GF) have to upgrade the fabs?

Upgrading fabs takes money in the bank, and thats something AMD doesn't have.

Where would AMD's fab process be if they were still the sole owner? how many qualcomm chips would AMD be producing?

The only thing hurting AMD in the short term is the excusivity to GF.

AMD had a five-year deal for the exclusive right to manufacture certain 28 nanometer Fusion APU chips, and the foundry is now waiving that right.

this is whats costing AMD 450M over a period of 2 years, to get rid of exclusivity.

 

[JAYDEEJOHN]

I never said GF was ahead, I said doing this allows for the monies, which Intel could only dream about, gives them such ability.
Besides Intel, GF has the resources to get to lead process, which may happen.

As to underperforming, thats AMDs fault, as to time to market, GFs and AMDs together, which wouldnt have changed hadnt the sell off occured.

Look at it this way, if AMD is more timely because of say TSMC, who wins?
If their product performs better because of say TSMC, again, who wins?

Also, Keller may not have come avoard if things werent different, and this is a win for team green.

Dont forget I also said if AMD deserves a hit theyll get one.
Well, BD was awful, a step back.
PD, a tad better, and what BD should have been.
So, this isnt GFs fault, or time to market etc, PD should have been here almost 2 years ago, and that lies on AMDs doorstep

 

[dozerman]

Too many people talking gaming and calling it a failure. The real question is in the server performance where the real money is.

 

[Cazalan]

I figured it had to be something like that. I was scratching my head trying to remember which new Opteron or Xeon slot processor I had missed.

The SeaMicro blades are similar but a bit larger than a P2.

They're 5x11" but with 8 physical CPUs.

 

[Cazalan]

as I stated, Intel is trying to force motherboard vendors into the likes of hp, dell, acer, ect.

motherboards are cheap to make in comparison to a full computer.

small computer repair shops will be hit the hardest ... "to fix your dead cpu, this will cost $500 for the motherboard with the soldered cpu" ... you can buy a new computer for that, customer leaves and goes to best buy.

The small computer repair shops have been dead for a while. The DIY market has been in rapid decline.

The gigabytes/asus/etc make the most money selling into the HP/Dell's. The PC box builders don't really make their own motherboards. They're just rebranded. Not much change there.

They're going to have to adjust. Intel wants to embed the system RAM next.

As more sectors move to SoC there's not much left to replace. Connectors are the weakest link right now. USB ports. HDMI ports. Those are the things getting torn up by too many insertions and ESD.

 

[earl45]

I never said GF was ahead, I said doing this allows for the monies, which Intel could only dream about, gives them such ability.
Besides Intel, GF has the resources to get to lead process, which may happen.

As to underperforming, thats AMDs fault, as to time to market, GFs and AMDs together, which wouldnt have changed hadnt the sell off occured.

Look at it this way, if AMD is more timely because of say TSMC, who wins?
If their product performs better because of say TSMC, again, who wins?

Also, Keller may not have come avoard if things werent different, and this is a win for team green.

Dont forget I also said if AMD deserves a hit theyll get one.
Well, BD was awful, a step back.
PD, a tad better, and what BD should have been.
So, this isnt GFs fault, or time to market etc, PD should have been here almost 2 years ago, and that lies on AMDs doorstep

Having all the money in the world don't make GF a better CPU company.
If it was all about money Apple would be one of the best CPU company's around.

 

[JAYDEEJOHN]

They have resources in IP, some great designers and engineers, but no monies, til the sale.
Now, newer minds can be added as well, plus all the best machinery.
This wouldnt have happened at an accelerated pace if AMD had held on.

 

[Cazalan]

If AMD kept the fab they would still be at 45nm. No money to invest in it. They had to sell.

 

[esrever]

AMD should have never had the deal with glofo. It hasn't paid off and its just been delay after delay from glofo and the processes never meeting expectations.

 

[Cazalan]

AMD should have never had the deal with glofo. It hasn't paid off and its just been delay after delay from glofo and the processes never meeting expectations.

They probably didn't have a choice in the initial deal. They had to sell that fab and it was a package deal. At the time they hadn't been profitable for a couple years, and the investors were willing to kick in 10+ billion to push the fab forward.

After the sale they eventually became profitable again.

I still don't see the big deal about GF. They yields have been good for a while. BD wasn't a flop because of GF. It's a weak design plain and simple. The chips are clocking just as high as Intel's 32nm process. Look at the Steamroller fixes and you'll see Bulldozer was crippled in the instruction decoder. That had nothing to do with GF.

The module design was a gamble that didn't pay off. AMD's best sales came with Llano and Bobcat which are both K10.5 derivatives. Now they're pushing Hondo SoC and Jaguar which are the same family. Single to quad cores that are independently clocked.

Even if AMD had tried to switch the CPUs to TSMC last year it wouldn't have mattered. TSMC didn't have the capacity to take it on.

 

[esrever]

the chips were designed to run higher. Just looking at the power consumption and the OC numbers. If you can cool bulldozer, it OCed. The leaking 32nm process made it impossible to run the chips at high speeds.

The initial 32nm process was garbage on llano and bulldozer. Yields were abysmal. Power consumption was high. Not to mention it was late to begin with.

The 45nm was pretty bad at the beginning also.

they still don't have the 28nm up after 2 years behind scheduled.

 

[de5_Roy]

a lot of people, not just in this site, seem to think amd was 'quick' to deliver piledriver after bulldozer came out 'so late'. since amd is dependent on glofo, i thought that both pd and bd are late, so was glofo with their fabrications.
old roadmap:

(shiny!) new roadmap:

i didn't iunderstand why the gaps became two years and why 32nm was skipped and why glofo lists 28nm still under 2011. if i read the old and new roadmaps right, we should be playing with 28nm steamroller apus since... q2-q3 2011...?

AFG: I hope you don't mind mate but I fixed your picture issue

 

[fazers_on_stun]

As to underperforming, thats AMDs fault, as to time to market, GFs and AMDs together, which wouldnt have changed hadnt the sell off occured.

Of course, that is speculation as AMD did spin off the fabs, due to having spent far too much on the ATI buyout and then Barcelona being a disaster.

IMO, the spinoff did nothing to promote cooperation between design & fab, and in fact almost assuredly hindered it due to each company having different goals and priorities. Certainly the cost savings of amortizing the SOI costs over multiple customers never materialized, plus GF switching to gate-last HKMG after being gate-first on 32nm didn't help matters either.

Dont forget I also said if AMD deserves a hit theyll get one.
Well, BD was awful, a step back.
PD, a tad better, and what BD should have been.
So, this isnt GFs fault, or time to market etc, PD should have been here almost 2 years ago, and that lies on AMDs doorstep

You forgot Llano also being delayed due to low yields.

Whether the problem was AMD not following the design rules for manufacturing properly, or the DRM being not limiting enough and causing problems, I don't think we know enough to assign blame completely one way or another. IMO, the truth probably lies in-between so both companies deserve blame.

 

[fazers_on_stun]

Too many people talking gaming and calling it a failure. The real question is in the server performance where the real money is.

Heh, we heard ya the first time ..

Anyway AMD has been stuck around the 5% mark in server marketshare for over a year now - if PD and/or the SeaMicro effort pays off, we might get a hint during the Q4 earnings report due out in about 6 or 7 weeks from now. But AMD undergoing another 15% layoff a few months after the last one, plus selling their Austin campus, plus seeking out Goldman Sachs and getting another cash bailout from ATIC does not bode well for any possible sudden increase in their cash flow from the most lucrative segment.

Only thing I've seen recently is Rory Read stating something about black Friday not sucking too much, which would seem to be about the consumer segment as I doubt IT buyers honor the black Friday tradition too much and line up at night to be first in line to buy a server farm at 50% off prices ..

 

[fazers_on_stun]

The small computer repair shops have been dead for a while. The DIY market has been in rapid decline.

The gigabytes/asus/etc make the most money selling into the HP/Dell's. The PC box builders don't really make their own motherboards. They're just rebranded. Not much change there.

They're going to have to adjust. Intel wants to embed the system RAM next.

As more sectors move to SoC there's not much left to replace. Connectors are the weakest link right now. USB ports. HDMI ports. Those are the things getting torn up by too many insertions and ESD.

Actually the one nearest my house in northern Virginia just expanded their floor space. I bought a couple storage drives from them recently, and they seem to always have a few customers. Of course, we are considered a somewhat affluent market so that could be it..