Samsung and GLOBALFOUNDRIES to Deliver Multi-Sourced 14 nm FinFET Offering

[Cristian_25H]

Samsung Electronics Co., Ltd. and GLOBALFOUNDRIES today announced a new strategic collaboration to deliver global capacity for 14 nanometer (nm) FinFET process technology. For the first time, the industry's most advanced 14 nm FinFET technology will be available at both Samsung and GLOBALFOUNDRIES, giving customers the assurance of supply that can only come from true design compatibility at multiple sources across the globe.

The new collaboration will leverage the companies' worldwide leading-edge semiconductor manufacturing capabilities, with volume production at Samsung's fabs in Hwaseong, Korea and Austin, Texas, as well as GLOBALFOUNDRIES' fab in Saratoga, New York.



Developed by Samsung and licensed to GLOBALFOUNDRIES, the 14 nm FinFET process is based on a technology platform that has already gained traction as the leading choice for high-volume, power-efficient system-on-chip (SoC) designs. The platform taps the benefits of three-dimensional, fully depleted FinFET transistors to overcome the limitations of planar transistor technology, enabling up to 20 percent higher speed, 35 percent less power and 15 percent area scaling over industry 20 nm planar technology.

The platform is the first FinFET technology in the foundry industry to provide true area scaling from 20 nm. The technology features a smaller contacted gate pitch for higher logic packing density and smaller SRAM bitcells to meet the increasing demand for memory content in advanced SoCs, while still leveraging the proven interconnect scheme from 20 nm to offer the benefits of FinFET technology with reduced risk and the fastest time-to-market.

Through this multi-year exclusive technology license, process design kits (PDKs) are available now, allowing customers to start designing with models, design rule manuals, and technology files that have been developed based on silicon results from 14nm FinFET test chips. Mass production for the 14 nm FinFET technology will begin at the end of 2014.

"This unprecedented collaboration will result in a global capacity footprint for 14 nm FinFET technology that provides AMD with enhanced capabilities to bring our innovative IP into silicon on leading-edge technologies," said Lisa Su, senior vice president and general manager of Global Business Units at AMD. "The work that GLOBALFOUNDRIES and Samsung are doing together will help AMD deliver our next generation of groundbreaking products with new levels of processing and graphics capabilities to devices ranging from low-power mobile devices, to next-generation dense servers to high-performance embedded solutions."

"This strategic collaboration extends the value proposition of a single GDSII multi-sourcing to the FinFET nodes. With this true multi-source platform, Samsung and GLOBALFOUNDRIES have made it easy for fabless semiconductor companies to access FinFET technology and increase first-time silicon success," said Dr. Stephen Woo, president of System LSI Business, Device Solutions Division, Samsung Electronics. "Through this collaboration, we are advancing the foundry business and support model to satisfy what customers have been asking for."

"Today's announcement is further proof of the importance of collaboration to enable continued innovation in semiconductor manufacturing," said GLOBALFOUNDRIES CEO Sanjay Jha. "With this industry-first alignment of 14 nm FinFET production capabilities, we can offer greater choice and flexibility to the world's leading fabless semiconductor companies, while helping the fabless industry to maintain its leadership in the mobile device market."

View at TechPowerUp Main Site

 

[Jorge]

This deal was forced by Apple because Samsung and GloFo will now produce their CPUs. Hopefully AMD will benefit from the tech and finally deliver a competitive discrete CPU line.

 

[Prima.Vera]

This deal was forced by Apple because Samsung and GloFo will now produce their CPUs. Hopefully AMD will benefit from the tech and finally deliver a competitive discrete CPU line.

 

[alwayssts]

LOL. Agreed.

But still, who knows what this could end up becoming. It makes sense the CPA would do something like this, especially in the wake of how IBM (the glue of the CPA) has been indicating they would like to have a smaller role and distance themselves from fabrication at some point (first with their fabs, then with r&d). Considering the fact they (Big Blue) championed both SOI and gate-first, both seemingly abandoned (or at least the former put in less priority in our increasing mobile/low-voltage/power/ARM-driven world) for the sake of manufacturing difficulties (or perhaps IBM not fully understanding their partners capabilities), perhaps it makes even more sense. It probably won't happen immediately, but ties like this are a step in the direction that one day the flock will be able to coordinate and function without their shepherd, which has to happen for us all to move forward and for them to realistically compete. Perhaps Samsung could take a more R&D heavy role while the whole of GF (and subsidiaries) could focus on manufacturing, or something to that effect. I hate to stereotype, but I could very much see that happening given the trend of Japanese and increasingly Korean innovation (R&D) through not only domestic but foreign hires to their companies, while the Chinese, Taiwanese, Singaporeans etc are more attune to huge production of known technology, and getting bigger by the day. Coordination of these strengths across multiple fields could see huge benefits to society, and I don't see why it couldn't happen here.

As for the 14nm process, who knows whom may use it? I won't jump to any conclusions, but it could be Apple. Heck, if they could swallow their pride (and stop worrying about IP getting stolen by AMD) even nvidia may use them. The 14nm process is the first time the CPA, or more notably GF, could actually prove themselves as a technology equal, if not superior to their competition (at least in some markets). While surely AMD could use the benefit given how far they've drifted behind Intel(/ARM-licensing companies), the company that really needs the win is GF if they want to be any kind of threat to TSMC. I hope this process is a good one, and that it comes up quickly so it can be a contemporary to what TSMC and Intel (at least at first) can offer. Given how difficult and likely long it will be between upcoming processes after this transition, we need all competition (in both manufacturing and end-products) we can get.

 

[john_]

Charlie at semi celebrates that he predicted this one. Also pointing at Apple as a first major customer. Who knows? Maybe. Going full speed to 14nm and partnering with one of your competitors for this it does mean that you have someone big waiting in the corner. You don't do it just for the chance to overshadow TSMC in the news.
Also AMD could use this tech for it's future chips, APUs(jaguar's successors will be great at 14nm) and GPUs. Nvidia might also prepare for this in case TSMC mess up with 14nm.

 

[alwayssts]

Charlie at semi celebrates that he predicted this one. Also pointing at Apple as a first major customer. Who knows? Maybe. Going full speed to 14nm and partnering with one of your competitors for this it does mean that you have someone big waiting in the corner. You don't do it just for the chance to overshadow TSMC in the news.
Also AMD could use this tech for it's future chips, APUs(jaguar's successors will be great at 14nm) and GPUs. Nvidia might also prepare for this in case TSMC mess up with 14nm.

Oh, I agree there is a reason. I just don't know if it's an Apple contract, which obviously would be anyone's first logical leap in assumption. This could be happening simply for the POTENTIAL of luring Apple. Obviously the main requirements of Apple are strict manufacturing practices and large quantities, and has shunned suppliers for both in the past (even during production). Samsung has the former, GF has the later. I'm simply saying there is potentially more to that mutually beneficial relationship than purely concerning Cupertino.

As for merits of the process, I have a feeling TSMC's 16nm (which is essentially the same thing) will still be the better tech, but probably not by a gigantic margin. That said, things like quantity, initial time to market (which may or may not beat tsmc given the historical accuracy of both roadmaps), and pricing could all be positively effected because of this arrangement, and could lead to some big contracts. The fact there is coordination is simply promising. Even if they don't have the better process (which is only speculative on my part), the fact ODMs can now distribute designs across two very good, if not similar-enough performig manufacturing chains only means more better products faster and lower prices.

Remember UMC old-timers? Can't help but wonder if the CPA will be leveraged as a new variation on how they were often used. Perhaps tsmc is faster, but cpa can make more product. Maybe the cpa comes first, and tsmc is used for more expensive products. Obviously in this case (since we know so little about how the processes are performing, let-alone their mass production schedules across fabs) those roles may switch, but there is some potential for interesting things to happen.

 

[Hilux SSRG]

Good for GlobalFoundries as they cancel their own 14nm development and good for Amd to stay in the game with Intel's 14nm tech.

 

[Jurassic1024]

This will push Intel even more on the mobile side. I'm okay with this.

 

[HumanSmoke]

Oh, I agree there is a reason. I just don't know if it's an Apple contract, which obviously would be anyone's first logical leap in assumption.

That should be an interesting dance given the current dance Cupertino and Samsung are involved in. AFAIK, GloFo is just licensing Samsung's foundry process because their own 14nm-XM was always going to be late to the party...and is now dead in the water. A far cry from a year ago.

If Apple and Samsung aren't buddies at the moment then it probably makes sense for GloFo to licence the fab process and court Apple, since Samsung seems to have a pretty full order book for their own fab capacity.

As for merits of the process, I have a feeling TSMC's 16nm (which is essentially the same thing) will still be the better tech, but probably not by a gigantic margin.

My thoughts exactly. TSMC if nothing else, has a vast knowledge of high transistor density, high power IC's - probably only second to Intel if Knights Landing is any indication. Samsung's (and now GloFo, as well as UMC) seem to be tailoring for the SoC market rather than high power/high density.