[ad_1]
The big picture: Samsung’s upcoming 3nm class lithography process will be the first to use gate-all-around transistor (GAA) technology. Faced with concerns raised about whether the company will have the GAA designs in time to keep pace with its competitor TSMC, it insists that its first iteration of the technology will begin mass production in 2022.
Samsung’s public roadmap was presented recently at its 2021 Foundry Forum in China. Samsung’s 3GAE (3nm, GAA-Early), the first iteration of their 3nm technology, was particularly noticeable in its absence on the slide shown. 3GAE was initially revealed in 2019 alongside its follow-up, 3GAP (3nm, GAA-Plus), but only the latter was featured in the presentation.
Before this roadmap was revealed, there had already been concerns about the health of the node; 3GAE was originally slated for risky production in late 2020 and volume production in 2021, but the company only registered its first 3nm test chips last month.
In addition, Dr Chidi Chidambaram, vice president of engineering at Qualcomm, Samsung Foundry’s largest third-party customer, was cited by SemiAnalysis as believing that GAA technology would not reach production until 2023-24, at a recent event hosted by Applied Materials.
While the estimate is likely intentionally vague for NDA reasons, it still puts Samsung a year behind TSMC’s plans for volume production of 3nm class silicon next year. This, combined with the absence of 3GAE on the published roadmap, led to speculation that the node had been completely ignored in favor of 3GAP.
Just to explain the name of Samsung: 3GAE (Gate-All-Around Early) is missing here. This indicates that the first process is not viable for mass production and has been completely killed. 3GAP (Gate-All-Around Plus) is now scheduled for HVM in 2023, which confirms Qualcomm’s statement https://t.co/9GeMcLq3jt
– Andreas Schilling (@aschiling) July 7, 2021
Contacted for comment by Anandtech, the company said 3GAE is still scheduled to go into mass production in 2022, and the node may not be released to the public simply because it is reserved for internal use at Samsung LSI.
However, although this was done with other variations of the company’s “-Early” process in the past, these were still marked on the map; 5LPE was used by third parties in Qualcomm’s Snapdragon 888, but this node was itself derived from 7LPP.
The company also offers a 4LPP, based on the more traditional FinFET technology. This one (and its predecessor 4LPE) are now listed as their own family of nodes, instead of being an evolution of the 5nm / 7nm class technologies.
It may be that the process has improved sufficiently substantial to warrant marketing as a new “title” node, or simply because of larger differences in design and manufacture.
Still, the lack of 3GAE on the public roadmap and the FinFET-based 4LPP node straddling Samsung’s stated 2022 target for 3GAE draws uncomfortable comparisons to Cannon Lake, Intel’s first 10nm processors.
Those technically shipped in 2018 to meet investor commitments, but were so bad that they were ignored in favor of further evolutions of the old 14nm node; the company now completely ignores this generation, insisting that Ice Lake was its first real Range of 10 nm.
But, perhaps, at least the oft-mocked 14 +++ was more easily understood than a branching tree of processes and their subevolutions.
[ad_2]
Source link