AMD moving from GloFo to TSMC with Zen

inb4 Fudzilla sucks or clickbait

Seriously though, if TSMC can do it better than GloFo, why the hell not? Especially if TSMC can help get Zen to market quicker.
 
inb4 Fudzilla sucks or clickbait

Seriously though, if TSMC can do it better than GloFo, why the hell not? Especially if TSMC can help get Zen to market quicker.

Well supposedly its been delayed to 2017. Not so quick
 
If Zen comes out in 2017 it will be competing with Cannon Lake. Maybe it'll be competitive, who knows. Intel has been sitting on their ass for 4 years.


Don't mind me, I'm annoyed because I want an NVMe SSD but I can't the purchase of a new CPU and board because they're not that much faster than my overclocked 2500+. Maybe if they put out a sub $600 8-core chip...
 
the good old days up udgrading every 12 to 18 months with much faster hardware is over... i still dont feel justified ugrading my IVY.... run everything i do great... though i want NVMe SSD too
 
the good old days up udgrading every 12 to 18 months with much faster hardware is over... i still dont feel justified ugrading my IVY.... run everything i do great... though i want NVMe SSD too

Sad when the next big upgrade is a SSD. :( I wish MS made a Windows 10 RISC Edition so we could have more options for processors since Intel and AMD have been incapable of innovating lately.
 
Sad when the next big upgrade is a SSD. :( I wish MS made a Windows 10 RISC Edition so we could have more options for processors since Intel and AMD have been incapable of innovating lately.

There's already a Windows 10 "RISC edition", its called Windows for ARM.
 
2017?! I thought it was due out around October 2016?

It IS still scheduled for a Q4 2016 launch according to recent statements from AMD.

Remember that not only is the "source" of this FUDzilla, but there are many FUD-ers in these very forums who's main purpose appears to be spreading their tainted non-information regarding AMD all over these forums and others.

Until AMD themselves indicate the product timeline for Zen has changed or that timeframe comes and goes without Zen's launch, don't put any weight into it.
 
so we'll finally get core2duo performance out of AMD?

Lol. It isn't that bad, but I understand the feeling. It does seem like eons ago that AMD was competitive.

I really hope Zen gets them back on track. However, I feel that realistically that even if Zen is a massive success that it will only stop the bleeding and not necessarily bring them back to profitability. AMD does have an unfortunate track record of bad decisions in the past 10 years. Fortunately, bringing back Jim Keller (albeit temporarily) and this possible rumor of switch to TSMC are definitely good signs of change.
 
so we'll finally get core2duo performance out of AMD?

Core2Duo was miles ahead of AMD at the time it was released, so that would be awesome for AMD to turn the tides like that. AMD is overdue for one of its classic Intel whippings.
Let's hop Zen brings back the tradition :D
 
I want something that is genuinely 40% faster per clock (vs fx) and has an upper clock speed in the 4-5ghz range without requiring 140w. But that'll just mean it's as good as intel, so they'll still have to sell cheaper. If they want to charge intel prices then they had better be genuinely better or pack in more cores or something.

Personally I wouldn't wait for it.
 
Core2Duo was miles ahead of AMD at the time it was released, so that would be awesome for AMD to turn the tides like that. AMD is overdue for one of its classic Intel whippings.
Let's hop Zen brings back the tradition :D

Indeed. Silence is good or bad. I'm very anxious to see what happens. I would love to see the kind of asswhoopin we saw with the Athlon 64, or the (ATi) Radeon 9700 Pro. It's well overdue.
 
Indeed. Silence is good or bad. I'm very anxious to see what happens. I would love to see the kind of asswhoopin we saw with the Athlon 64, or the (ATi) Radeon 9700 Pro. It's well overdue.

At the very least it would be interesting to see just how much R&D Intel is sitting one while they putt along with their 5% speed increments.
 
At the very least it would be interesting to see just how much R&D Intel is sitting one while they putt along with their 5% speed increments.



I say the possibility is out there that they're hitting some limits as well. Their CPU marketing is as bad as AMD's GPU marketing. ..
 
If they would just not dedicate half of the die for a GPU and use that space for more CPU then maybe they can actually make a processor worth buying.

Throw in one of those 17mm2 ARM-style GPUs to handle video display and decoding and be done with it.
 
Zen didn't get delayed to 2017. As usual that bit of "information" came from morons misreading comments that came from an opinion from some guy who wrote an article about AMD recently on a non-tech website of all things. AMD has said that they will launch Zen and then get the first year of full revenue in 2017. You can't get a full year of revenue in 2017 if the product launches in 2017. My guess is that we get a launch around this time next year with higher availability in early 2017.

As for GloFo they have been responsible for many slip-ups and botched launches for AMD for years now. Llano was supposed to come out in Q1 2011 but got delayed three fucking quarters because of them. Not to mentio Llano was supposed to achieve much higher clocks than it actually wound up achieving. Similar issues with the initial 32nm SOI with the original Bulldozer launch.
 
If they would just not dedicate half of the die for a GPU and use that space for more CPU then maybe they can actually make a processor worth buying.

Throw in one of those 17mm2 ARM-style GPUs to handle video display and decoding and be done with it.

It's been known for a long time now that the initial Zen processors will be purely CPU's. The APU variants launch later in 2017. The iGPU has no impact on the x86 cores' performance anyway. People don't seem to mind paying $300+ for an Intel quad-core with SMT and over half the die being graphics IP.
 
http://www.kitguru.net/components/anton-shilov/globalfoundries-14nm-yields-are-exceeding-our-plans/

"Our 14nm FinFET ramp is exceeding plan with best-in-class yield and defect density,” said Jason Gorss, senior manager of corporate and technology communications at GlobalFoundries. “The early-access version of the technology (14LPE) was qualified in January and is well on its way to volume production, meeting yield targets on lead customer products.”

http://www.kitguru.net/components/a...cts-using-second-gen-14nm-process-technology/

"“The performance-enhanced version of the technology (14LPP) is set for qualification in the second half of 2015, with the volume ramp beginning in early 2016,” said Jason Gorss, senior manager of corporate and technology communications at GlobalFoundries."

"“Prototyping on test vehicles has demonstrated excellent logic and SRAM yields and performance at near 100% of target,” explained the official for GlobalFoundries."
 
http://www.kitguru.net/components/anton-shilov/globalfoundries-14nm-yields-are-exceeding-our-plans/

"Our 14nm FinFET ramp is exceeding plan with best-in-class yield and defect density,” said Jason Gorss, senior manager of corporate and technology communications at GlobalFoundries. “The early-access version of the technology (14LPE) was qualified in January and is well on its way to volume production, meeting yield targets on lead customer products.”

http://www.kitguru.net/components/a...cts-using-second-gen-14nm-process-technology/

"“The performance-enhanced version of the technology (14LPP) is set for qualification in the second half of 2015, with the volume ramp beginning in early 2016,” said Jason Gorss, senior manager of corporate and technology communications at GlobalFoundries."

"“Prototyping on test vehicles has demonstrated excellent logic and SRAM yields and performance at near 100% of target,” explained the official for GlobalFoundries."

Somebody isn't telling the truth then
 
any else concerned we have no large state of the art ic fabs in the US...
 
If I'd have to guess: Zen CPUs will be manufactured by TSMC, Zen APUs by GloFo.
 
Is this an advanced stage of rumour mill ping pong that sites tend to play ?

yes, 18 months later they can always point to one of the articles and say; "look, we were right", safe in the knowledge that the opposite opinion piece is long lost from sight to the casual search.
 
You are not the only one.

There is almost no reason for Samsung to purchase AMD. The x86 license is non-transferable, and Samsung has no interest in getting into the Desktop market anyway. All off the low-power tech ATI used to own was sold to Qualcomm following it's purchase by AMD, so there's little reason for Samsung to purchase them for their GPU tech either. The only attractive thing for Samsung would be the patent portfolio, but I can't think of a single one that Samsung really needs.

Nevermind the debt and profit losses Samsung would inherit for no financial gain.

So yeah, not happening.
 
Nevermind the debt and profit losses Samsung would inherit for no financial gain.
So yeah, not happening.

The thing is people are to much like sheep on this subject. The whole thing gets a circle jerk and from time to time Forbes or any other financial based information site gets this "scoop" while it is not a scoop it is just something to generate traffic.

AMD can not be bought without killing it as it is now. Because Intel is not going to say yes to the transfer of the x86 license to a company with enough money to spend to make life hard on Intel.....
 
The thing is people are to much like sheep on this subject. The whole thing gets a circle jerk and from time to time Forbes or any other financial based information site gets this "scoop" while it is not a scoop it is just something to generate traffic.

AMD can not be bought without killing it as it is now. Because Intel is not going to say yes to the transfer of the x86 license to a company with enough money to spend to make life hard on Intel.....

SEC could.... also there are other options like AMD going private and Samsung taking a controlling share etc
 
AMD will not be using TSMC for their cpus.

I don't even know how this rumor started. They already announced that 14nm LPP was the node they were going to use. Which is interesting because this will be the first high performance design on this process. Fmax of transistors has been going down since 32 nm SOI. ( main reason Ivy Bridge/Haswelll/Skylake ) don't overclock quite as well as Sandy Bridge did or heck even Piledriver.


Most people associate node shrinks with improved performance. However these days it really only improves two metrics. reduced die space, and performance per watt. AMD with Zen will see reduced clock speeds, much better power consumption, and smaller dies. How much of a clock speed hit they will take is well pseudoscience. Biggest question here is can AMD actually improve IPC enough with Zen where they can tolerate reduced clock speeds.
 
Even though the article mentions that Jim Keller still works for AMD?

Just saying, the information seems out of date; rehashed article maybe?

The only mention of keller is in this sentence and simply says he was the lead developer of zen which has been confirmed for some time.
A big chunk of AMD's R&D budget is allocated to getting the "Zen" architecture right, with key stages of its development being handled by Jim Keller, the brains behind some of AMD's most commercially successful CPU cores.
 
Back
Top