Process Technology Overview
We have been very spoiled throughout the years. We likely did not realize exactly how spoiled we were until it became very obvious that the rate of process technology advances hit a virtual brick wall. Every 18 to 24 months we were treated to a new, faster, more efficient process node that was opened up to fabless semiconductor firms and we were treated to a new generation of products that would blow our hair back. Now we have been in a virtual standstill when it comes to new process nodes from the pure-play foundries.
Few expected the 28 nm node to live nearly as long as it has. Some of the first cracks in the façade actually came from Intel. Their 22 nm Tri-Gate (FinFET) process took a little bit longer to get off the ground than expected. We also noticed some interesting electrical features from the products developed on that process. Intel skewed away from higher clockspeeds and focused on efficiency and architectural improvements rather than staying at generally acceptable TDPs and leapfrogging the competition by clockspeed alone. Overclockers noticed that the newer parts did not reach the same clockspeed heights as previous products such as the 32 nm based Sandy Bridge processors. Whether this decision was intentional from Intel or not is debatable, but my gut feeling here is that they responded to the technical limitations of their 22 nm process. Yields and bins likely dictated the max clockspeeds attained on these new products. So instead of vaulting over AMD’s products, they just slowly started walking away from them.
Samsung is one of the first pure-play foundries to offer a working sub-20 nm FinFET product line. (Photo courtesy of ExtremeTech)
When 28 nm was released the plans on the books were to transition to 20 nm products based on planar transistors, thereby bypassing the added expense of developing FinFETs. It was widely expected that FinFETs were not necessarily required to address the needs of the market. Sadly, that did not turn out to be the case. There are many other factors as to why 20 nm planar parts are not common, but the limitations of that particular process node has made it a relatively niche process node that is appropriate for smaller, low power ASICs (like the latest Apple SOCs). The Apple A8 is rumored to be around 90 mm square, which is a far cry from the traditional midrange GPU that goes from 250 mm sq. to 400+ mm sq.
The essential difficulty of the 20 nm planar node appears to be a lack of power scaling to match the increased transistor density. TSMC and others have successfully packed in more transistors into every square mm as compared to 28 nm, but the electrical characteristics did not scale proportionally well. Yes, there are improvements there per transistor, but when designers pack in all those transistors into a large design, TDP and voltage issues start to arise. As TDP increases, it takes more power to drive the processor, which then leads to more heat. The GPU guys probably looked at this and figured out that while they can achieve a higher transistor density and a wider design, they will have to downclock the entire GPU to hit reasonable TDP levels. When adding these concerns to yields and bins for the new process, the advantages of going to 20 nm would be slim to none at the end of the day.
Subject: General Tech | December 11, 2014 - 12:50 PM | Jeremy Hellstrom
As best we know TSMC is the sole fabricator of Apple's A8 chips on 20nm process, but so far from what DigiTimes has been able to determine that is not the case for the upcoming A9 chips. TSMC plans to keep pricing the same as they move to 14nm process tech but both Samsung and GLOBALFOUNDRIES are in a position where they could decide to drop their pricing in order to win business. Qualcomm has already placed orders for its 14nm chips with TSMC and Samsung but it is possible that with the experience GLOBALFOUNDRIES has with the 14nm process thanks to business from AMD they may also be able to undercut TSMC's pricing, assuming their yields can stay up.
"Globalfoundries is striving to be among the major contract chipmakers of Qualcomm and Apple, vying for 14nm chip orders from the two vendors, according to industry sources."
Here is some more Tech News from around the web:
- Microsoft's dodgy new Exchange 2010 update breaks Outlook clients @ The Register
- 'Critical' security bugs dating back to 1987 found in X Window @ The Register
- Bellard Creates New Image Format To Replace JPEG @ Slashdot
- Microsoft lets YOU kill POODLE in Protected Mode sites @ The Register
- The TR Podcast 167: The best of 2014 & 2015, plus two freaking petabytes
Subject: General Tech | February 7, 2014 - 01:18 PM | Scott Michaud
Tagged: TSMC, IBM, GLOBALFOUNDRIES
Well this is something which I expect they will not sell to Lenovo...
IBM, one of the world's most advanced chip fabrication companies with the capability to manufacture on a 22nm node, is looking to sell this division. According to The Financial Times, via Ars Technica, the company selected Goldman Sachs to seek options. They are primarily looking for interested buyers but would also consider finding a business partner to offload the division into a joint venture.
The two initial candidates are GLOBALFOUNDRIES and TSMC.
Image Credit: IBM via ZDNet (Outside photographers are not allowed inside their fab lab).
IBM is not willing to get rid of its chip design ability. IBM creates many chips, often based on its own "Power Architecture". This trademark comes with their RISC-based instruction sets which rival ARM and x86. It forms the basis of the Xbox 360, the Cell processor found in the PS3 (and rarely elsewhere), and the last three Nintendo game consoles starting with the Gamecube.
Despite designing all of the above chips, only some were actually fabricated by IBM.
Personally, I am not sure how serious the earlier mentioned potential buyers are. It could have easily been someone who looked at the list of leading foundries and picked the top two. TSMC is not even a member of "the Common Platform" alliance, not to mention how small IBM is compared to them, so I cannot see much reason for TSMC to bother.
GLOBALFOUNDRIES is a different story, It would make sense for them to want that part of IBM (Josh notes they even share some resource centers). Still, the both of us wondered if they could afford the deal. ATIC, parent company of GLOBALFOUNDRIES, might be able to get the money from somewhere - but would they? They purchased Charter only just recently. Now, if they simply enter a partnership with IBM, that might be a different story than an outright purchase.
Fabrication is hard and expensive. Creating a foundry is about $10 billion, give or take a few billion depending on yield, and changing your equipment for new nodes or wafer sizes is not much cheaper. I can see IBM, a company that is increasing concerned with high profitability, wanting to let someone else deal with at least some of the volatility.
IBM has not commented on this rumor.
Subject: General Tech | January 10, 2014 - 12:18 PM | Jeremy Hellstrom
Tagged: UMC, SoFIA, Intel, GLOBALFOUNDRIES, atom, 28nm
GLOBALFOUNDRIES will be the primary supplier of Intel's 28nm baseband chips according to this unconfirmed report at DigiTimes. It seems that Intel really is moving towards a new business model and will be outsourcing some of their upcoming chips to both GLOFO and UMC. Their 28nm PolySiON process will be used to make the next generation of baseband transmitter chips and the new Atom SoC for cellphones and phablets will use TSMC's 28nm HKMG process. The higher end Broxton SoCs will remain at Intel and use their FinFET process. This is a big win for GLOFO and could mean the beginning of a lasting partnership with what was once an AMD asset.
"Intel has contracted Taiwan Semiconductor Manufacturing Company (TSMC) to manufacture its forthcoming Atom mobile processor series codenamed SoFIA, and also placed orders for entry-level baseband chips with Globalfoundries and United Microelectronics (UMC), according to industry sources."
Here is some more Tech News from around the web:
- Worldwide PC shipments decline 6.9% in 4Q13, says Gartner @ DigiTimes
- Microsoft is expected to put the Nvidia Tegra K1 in Surface tablets @ The Inquirer
- 7 Great New Open Source Projects @ Linux.com
- Intel launches 22nm Xeon E5-2400 v2 chips for servers @ The Inquirer
- Fairfax shovels another $250m into buying BlackBerry debt @ The Register
- Rogue Android: We show you how BlackBerry's pain can be your gain @ The Register
- Ford Exec: 'We Know Everyone Who Breaks the Law' Thanks To Our GPS In Your Car @ Slashdot
Subject: Editorial, General Tech | December 8, 2013 - 04:11 AM | Scott Michaud
Tagged: TSMC, GLOBALFOUNDRIES, broadcom
Josh Walrath titled the intro of his "Next Gen Graphics and Process Migration: 20nm and Beyond" editorial: "The Really Good Times are Over". Moore's Law predicts that, with each ~2 year generation, we will be able to double the transistor count of our integrated circuits. It does not, however, set a price.
A look into GlobalFoundries.
"Moore's Law is expensive" remarked Tom Kilroy during his Computex 2013 keynote. Intel spends about $12 billion USD in capital, every year, to keep the transistors coming. It shows. They are significantly ahead of their peers in terms of process technology. Intel is a very profitable company who can squirrel away justifications for these research and development expenses across numerous products and services.
The benefits of a process shrink are typically three-fold: increased performance, decreased power consumption, and lower cost per chip (as a single wafer is better utilized). Chairman and CTO of Broadcom, Henry Samueli, told reporters that manufacturing complexity is pushing chip developers into a situation where one of those three benefits must be sacrificed for the other two.
You are suddenly no longer searching for an overall better solution. You are searching for a more optimized solution in many respects but with inherent tradeoffs.
He expects GlobalFoundries and TSMC to catch up to Intel and "the cost curve should come back to normal". Still, he sees another wall coming up when we hit the 5nm point (you can count the width or height of these transistors, in atoms, using two hands) and even more problems beyond that.
Image Credit: IONAS
From my perspective: at some point, we will need to say goodbye to electronic integrated circuits. The theorists are already working on how we can develop integrated circuits using non-electronic materials. For instance, during the end of my Physics undergraduate degree, my thesis adviser was working on nonlinear optics within photonic crystals; waveguides which transmit optical frequency light rather than radio frequency electric waves. Of course I do not believe his research was on Optical Integrated Circuits, but that is not really the point.
Humanity is great at solving problems when backs are against walls. But, what problem will they try?
Power consumption? Cost? Performance?
Subject: General Tech | November 18, 2013 - 12:44 PM | Jeremy Hellstrom
Tagged: qualcomm, GLOBALFOUNDRIES, Samsung, TSMC
Qualcomm is looking to diversify their supply chain and move away from dependence on TSMC and their 28nm node. They have some qualifications for their suitor to meet and being one of the larger customers means that they just might get it. Their requirement is for the rumoured Samsung and GLOBALFOUNDRIES partnership to become stable and for Samsung to use GF as a sub-contractor to make chips for Apple. If you believe all the hints we are getting the partnership could grow and it would give Qualcomm a supplier who is financially stable and still has enough free resources to fab Qualcomm's chips in the desired volume. This is the news out of DigiTimes this morning.
"Qualcomm reportedly hopes Samsung Electronics and Globalfoundries can form an alliance, as the fabless IC vendor seeks to reduce its reliance on Taiwan Semiconductor Manufacturing Company (TSMC) for its advanced chips, according to industry sources."
Here is some more Tech News from around the web:
- Linux backdoor squirts code into SSH to keep its badness buried @ The Register
- Microsoft launches a 3D printing app for Windows 8.1 @ The Inquirer
- 'Planned maintenance' CRIPPLES nearly HALF of all Salesforce instances in Europe, US @ The Register
- 'I'm BIG, I'm BALD and I'm LOUD!' Blubbering Ballmer admits HE was Microsoft's problem @ The Register
- Awesome BlizzCon 2013 CosPlay Pictures By Legit Reviews
- ASUS RT-AC56U Dual Band Wireless-AC1200 Gigabit Router Review @HiTech Legion
The Really Good Times are Over
We really do not realize how good we had it. Sure, we could apply that to budget surpluses and the time before the rise of global terrorism, but in this case I am talking about the predictable advancement of graphics due to both design expertise and improvements in process technology. Moore’s law has been exceptionally kind to graphics. We can look back and when we plot the course of these graphics companies, they have actually outstripped Moore in terms of transistor density from generation to generation. Most of this is due to better tools and the expertise gained in what is still a fairly new endeavor as compared to CPUs (the first true 3D accelerators were released in the 1993/94 timeframe).
The complexity of a modern 3D chip is truly mind-boggling. To get a good idea of where we came from, we must look back at the first generations of products that we could actually purchase. The original 3Dfx Voodoo Graphics was comprised of a raster chip and a texture chip, each contained approximately 1 million transistors (give or take) and were made on a then available .5 micron process (we shall call it 500 nm from here on out to give a sense of perspective with modern process technology). The chips were clocked between 47 and 50 MHz (though often could be clocked up to 57 MHz by going into the init file and putting in “SET SST_GRXCLK=57”… btw, SST stood for Sellers/Smith/Tarolli, the founders of 3Dfx). This revolutionary graphics card at the time could push out 47 to 50 megapixels and had 4 MB of VRAM and was released in the beginning of 1996.
My first 3D graphics card was the Orchid Righteous 3D. Voodoo Graphics was really the first successful consumer 3D graphics card. Yes, there were others before it, but Voodoo Graphics had the largest impact of them all.
In 1998 3Dfx released the Voodoo 2, and it was a significant jump in complexity from the original. These chips were fabricated on a 350 nm process. There were three chips to each card, one of which was the raster chip and the other two were texture chips. At the top end of the product stack was the 12 MB cards. The raster chip had 4 MB of VRAM available to it while each texture chip had 4 MB of VRAM for texture storage. Not only did this product double performance from the Voodoo Graphics, it was able to run in single card configurations at 800x600 (as compared to the max 640x480 of the Voodoo Graphics). This is the same time as when NVIDIA started to become a very aggressive competitor with the Riva TnT and ATI was about to ship the Rage 128.
Subject: General Tech | April 18, 2013 - 01:46 PM | Ken Addison
Tagged: vsync, vertex 3.20, podcast, pcper, overclocking, ocz, haswell, gtx 780, GLOBALFOUNDRIES, gigabyte brix, frame rating
PC Perspective Podcast #247 - 04/18/2013
Join us this week as we discuss Frame Rating and Vsync, the future of GLOBALFOUNDRIES, the OCZ Vertex 3.20 and more!
The URL for the podcast is: http://pcper.com/podcast - Share with your friends!
- iTunes - Subscribe to the podcast directly through the Store
- RSS - Subscribe through your regular RSS reader
- MP3 - Direct download link to the MP3 file
Hosts: Ryan Shrout, Jeremy Hellstrom, Josh Walrath, and Allyn Malventano
Program length: 1:07:41
0:01:08 Win the Roccat ISKU Keyboard
Week in Review:
News items of interest:
Jeremy: support Full Control not just because they're nordic
Allyn: (portable headsets that don't suck)
1-888-38-PCPER or firstname.lastname@example.org
Taking a Fresh Look at GLOBALFOUNDRIES
It has been a while since we last talked about GLOBALFOUNDRIES, and it is high time to do so. So why the long wait between updates? Well, I think the long and short of it is a lack of execution from their stated roadmaps from around 2009 on. When GF first came on the scene they had a very aggressive roadmap about where their process technology will be and how it will be implemented. I believe that GF first mentioned a working 28 nm process in a early 2011 timeframe. There was a lot of excitement in some corners as people expected next generation GPUs to be available around then using that process node.
Fab 1 is the facility where all 32 nm SOI and most 28 nm HKMG are produced.
Obviously GF did not get that particular process up and running as expected. In fact, they had some real issues getting 32 nm SOI running in a timely manner. Llano was the first product GF produced on that particular node, as well as plenty of test wafers of Bulldozer parts. Both were delayed from when they were initially expected to hit, and both had fabrication issues. Time and money can fix most things when it comes to process technology, and eventually GF was able to solve what issues they had on their end. 32 nm SOI/HKMG is producing like gangbusters. AMD has improved their designs on their end to make things a bit easier as well at GF.
While shoring up the 32 nm process was of extreme importance to GF, it seemingly took resources away from further developing 28 nm and below processes. While work was still being done on these products, the roadmap was far too aggressive for what they were able to accomplish. The hits just kept coming though. AMD cut back on 32nm orders, which had a financial impact on both companies. It was cheaper for AMD to renegotiate the contract and take a penalty rather than order chips that it simply could not sell. GF then had lots of line space open on 32 nm SOI (Dresden) that could not be filled. AMD then voided another contract in which they suffered a larger penalty by opting to potentially utilize a second source for 28 nm HKMG production of their CPUs and APUs. AMD obviously was very uncomfortable about where GF was with their 28 nm process.
During all of this time GF was working to get their Luther Forest FAB 8 up and running. Building a new FAB is no small task. This is a multi-billion dollar endeavor and any new FAB design will have complications. Happily for GF, the development of this FAB has gone along seemingly according to plan. The FAB has achieved every major milestone in construction and deployment. Still, the risks involved with a FAB that could reach around $8 billion+ are immense.
2012 was not exactly the year that GF expected, or hoped for. It was tough on them and their partners. They also had more expenses such as acquiring Chartered back in 2009 and then acquiring the rather significant stake that AMD had in the company in the first place. During this time ATIC has been pumping money into GF to keep it afloat as well as its aspirations at being a major player in the fabrication industry.