GlobalFoundries Establishes Avera Semiconductor: a Custom Chip Company
by Anton Shilov on November 2, 2018 8:00 AM EST
GlobalFoundries this week announced that it has spun off its ASIC Solutions division, establishing Avera Semiconductor, a wholly owned subsidiary that will help fabless chip developers to design their products. Avera will work closely with GlobalFoundries' customers to enable them take advantage of various process technologies that GF has, but the company will also establish ties with other contract makers of semiconductors to help their clients develop chips to be made using leading edge process technologies at 7 nm and beyond.
Avera Semi will inherit a rich portfolio of silicon-proven IP containing Arm cores, performance and density-optimized SRAMs, embedded TCAMs, high-speed SerDes, interfaces, and other necessary things. In addition, Avera will offer production-proven design methodologies as well as advanced packaging options. Right now, Avera naturally has silicon-proven IP for GlobalFoundries’ manufacturing technologies as well as Samsung Foundry’s 14LPP fabrication process, but over time the company will have to gain silicon-proven IP for other contract makers of semiconductors, namely TSMC.
GlobalFoundries says that Avera’s team can serve clients from virtually every semiconductor industry, including AI/ML (think leading edge technologies), aerospace & defense (think special-purpose technologies as well as radiation-resistant semiconductors), HPC, storage, wired/wireless networking, and so on.
The new wholly owned subsidiary of GlobalFoundries has over 850 employees, an annual revenue of over $500 million, and ongoing projects worth $3 billion. By working not only with clients of GlobalFoundries, but expanding to customers of companies like Samsung Foundry and TSMC, Avera has a chance to increase its earnings over time. Avera Semi is led by Kevin O’Buckley, a former head of ASIC Solutions, who joined GlobalFoundries from IBM.
Avera Semiconductor will continue to work with a variety of industry partners, including Arm and Synopsys, who have already endorsed the new business unit.
Related Reading:
- GlobalFoundries Stops All 7nm Development: Opts To Focus on Specialized Processes
- Change of Strategy: A New GlobalFoundries CEO in Dr. Thomas Caulfield
- The Future of Silicon: An Exclusive Interview with Dr. Gary Patton, CTO of GlobalFoundries
- GlobalFoundries Announces 22FDX Milestone: $2 Billion in Design Wins
- GlobalFoundries to Expand Capacities, Build a Fab in China
Source: GlobalFoundries
Facebook
Twitter
RSS
11 Comments
View All Comments
wumpus - Friday, November 2, 2018 - link
Any idea why they are doing this? Investors demand "something", so they start rearranging deck chairs? I'd assume they expect Avera to make even less money than GF (or perhaps more, and possibly inflate the stock price...).If I wanted to fab a chip with GF, I'd want my tools from as close to the fab as possible, and this isn't helping. Maybe they think they can get tools from Samsung easier with a little distance between the tool company and the fab, but I can't imagine that really matters.
wumpus - Friday, November 2, 2018 - link
[hopefully this edits and not doubleposts]Any idea why they are doing this? Investors demand "something", so they start rearranging deck chairs? I'd assume they expect Avera to make even less money than GF (or perhaps more, and possibly inflate the stock price...).
If I wanted to fab a chip with GF, I'd want my tools from as close to the fab as possible, and this isn't helping. Maybe they think they can get tools from Samsung easier with a little distance between the tool company and the fab, but I can't imagine that really matters.
I guess the thing I missed is real deprication vs. tax deprication. I'm guessing fabs depricate in really nasty ways (although the outside building is a tax benefit, the insides only help if they suckered some locals into large corporate welfare). There might be some tax benefits here, but I'd still expect the real reason is to give the investors some indication that they are "doing something".
flgt - Friday, November 2, 2018 - link
Sounded like the ASIC design group had to be able to offer TSMC 7 nm (or whoever the latest process leader is) to be competitive in that business segment. It probably helped create clear IP segregation between the design house, TSMC fab, and GF fab that allows them to have access to TSMC's design libraries.wassick - Friday, November 2, 2018 - link
The press release indicated that they've already established a foundry partnership ("...., including a newly established foundry partnership on 7nm") -- so I'm wondering if there's any word on who've their starting with....frenchy_2001 - Friday, November 2, 2018 - link
It's not as if they have much choice, it can only be TSMC or Samsung.Historically, GF has been closer to Samsung (they shared some research and the resulting processes), but either, or even both, are possibilities.
DanNeely - Friday, November 2, 2018 - link
GloFo is abandoning development of new high performance processes. In the future they'll only be doing new RF and other specialty types of work. That means they won't be able to present themselves to end users as a 1 stop shop, which would leave them at a major disavantage vs competitors. Moving their consulting services away from the main company (if only slightly at the moment) will make it easier for the consulting group to work with TSMC/Samsung for products that GloFo doesn't have the ability to produce. I wouldn't be surprised if they end up having to fully release the consulting business to lower barriers to TSMC/Samsung more, but as GloFo itself they probably have 0 traction due to the latter needing to protect their IP.ibnmadhi - Saturday, November 3, 2018 - link
Might have something to do with their contract to manufacture CPUs for IBM... I suppose this allows for the CPUs to actually be fabbed at TSMC, but go through this company.peevee - Sunday, November 4, 2018 - link
It is a design house, not manufacturing house.peevee - Sunday, November 4, 2018 - link
"Any idea why they are doing this?"Because GF failed to implement "7" and even "10nm". So significant part of the market are not going to produce with GF. This way they at least hope to capture some consulting revenue, although it would be hard to compete with people who already have experience with TSMC and Samsung.
Unraveling of the terrible decisions which led to abandoning any further technological development...
del42sa - Monday, November 5, 2018 - link
Thy doing it to keep milking their customers like AMD via WSA....