CEA-Leti today announced that it has joined the GLOBALSOLUTIONS ecosystem as an ASIC provider, specifically to support GLOBALFOUNDRIES’ 22FDX (TM) technology platform.
Launched this summer, GLOBALFOUNDRIES’ 22FDX technology platform is the industry’s first 22nm FD-SOI semiconductor technology developed specifically to meet the ultra-low-power requirements of the next generation of connected devices. The versatility of the 22FDX platform is a result of unmatched design flexibility and intelligence, including software-controlled transistor body-biasing that provides real-time trade-offs between power and performance. Delivering FinFET-like performance and energy-efficiency at a cost comparable to 28nm planar technologies, the platform enables a new level of innovations on next-generation chips and sets new standards in-terms of user experience for Internet of Things (IoT), mainstream mobile, RF, and networking applications.
GLOBALSOLUTIONS was created more than five years ago to spur innovation in the semiconductor industry and assure chip designers receive world-class service from design conception to production. The ecosystem combines GLOBALSOLUTIONS’ internal resources with a broad spectrum of partners to efficiently enable the fastest time-to-volume for foundry customers.
“Together with our design services partners, we are able to offer a full suite of services and comprehensive turnkey solutions that confirms GLOBALFOUNDRIES’ leadership in providing high-performance customized products in the FD-SOI and ASIC markets,” said Gary Patton, chief technology officer and head of worldwide R&D at GLOBALFOUNDRIES. “Our expanded partnership with Leti further reflects our commitment to find design implementations that will accelerate time-to-volume and deliver ultra-low-power solutions to our customers.”
Earlier this year, Leti assigned a team of experts to GLOBALFOUNDRIES’ Dresden, Germany, Fab 1 to support ramp up of the platform. As an ecosystem partner, Leti will provide GLOBALFOUNDRIES’ customers circuit-design IP, including for its back-bias feature for FD-SOI, which enables exceptional performance at very low voltages with low leakage.
“This strategic partnership with GLOBALFOUNDRIES positions Leti to help a broad range of designers utilize FD-SOI technology’s significant strengths in ultra-low-power and high performance in their IoT and mobile devices with 22nm technology,” said Marie Semeria, Leti CEO. “In addition, it gives both sides’ customers increased access to our respective technologies. This kind of partnership is a key part of Leti’s global strategy.”
Sang Kim
Here are my comments. This is not the first time that 28FDX(28nm FDSOI) is supported by GF, Leti and CEA in the past but didn’t succeed. First, the 28nm bulk is in mass production for several years by major semiconductor companies but not 28FDX yet. “Delivering FinFET- like performance and …a cost comparable to 28nm planner technologies, ….etc”. In my opinion these statements are over exaggeration because as I said above, 28FDX is not manufactured yet because the most critical issue with 28FDX is unable to implement LDD(Lightly Doped Drain) designed to minimize the hot carrier generation as is done in 28nm bulk. Furthermore, the hot carrier generation becomes worse in 22FDX and beyond. Therefore, unless the hot carrier issue were resolved, 28FDX and beyond wouldn’t be manufactured. Since LDD can’t be implemented in 28FDX and beyond, 28FDX and beyond wouldn’t be manufacturable due to hot carrier reliability, therefore the end of FDX! Other critical issue for 22FDX and beyond is unable to suppress the leakage currents due to the short channel effects resulting in device failures. These are not only the end of 28FDX and 22FDX but also the end of FDX! That is why Intel choused, developed and manufactured FinFET device at the 22nm node to suppress the leakage current due to the short channel effects. The beauty of FinFET is that as long as FinW(width) is equal to or smaller than Lg(gate length or channel length), the leakage current is suppressed. That is why Intel 14nm FinFET is volume manufactured over two years and 10nm FinFET will be possibly in 2016 or this year. TSMC 16FDX will be manufactured also this year. How thin SOI layer will be required for 14FDX and 10FDX? About 4nm for 14FDX and about 3nm for 10 FDX respectively. Depositing such ultrathin 4nm and 3nm uniformly and reliably across 12″ wafers in the manufacturing line is extremely difficult or not manufacturable. If not manufacturable, the end of the FDX or FDSOI!
Sorry for one typo error correction: TSMC 16nmFDX from 5th line from the bottom should be TSMC 16nm FinFET instead.
HCI is not a problem. Leakage also not a problem in FDSOI. I’m sure that 22FDX will succdeed.
Sang Kim
I would like to add one key additional point on the hot carrier reliability that I have neglected to point out clearly regarding the 22nm FDX device especially. When the hot carriers(electrons and holes) are generated when the short channel 22nm FDX device is on, electrons drift to the drain, but where the holes to go? The holes have no where else to go but all go to the source through the un-doped SOI channel, forming a parasitic N+(source)/holes((+) diode at the source, resulting in large diode leakage current failures at the source. Therefore, FDX is not manufacture-able. These points were previously pointed out previously by me in EETimes.