IEDM 2017 + ISSCC 2018: Intel’s 10nm, switching to cobalt interconnects


Intel has scaled the SRAM at 0.62x for all the cells. The various cells are optimized for density, power, and performance. The high-density (PU:PG:PD = 1:1:1) cell has shrunk from 0.0499 squared micron down to 0.0312 while the high-performance (PU:PG:PD = 1:2:2) of 0.441 squared micron.

Intel’s 10nm SRAM Cells
Cell 14 nm 10 nm Scaling
High-Density 0.0499 µm² 0.0312 µm² 0.63x
High-Performance 0.0706 µm² 0.0441 µm² 0.63x
Low-Voltage 0.0588 µm² 0.0367 µm² 0.63x

We have plotted Intel’s 10nm SRAM cell on the graph along with all of their historical SRAM cell sizes.


Intel’s Historical SRAM trend.

Despite still being on Moore’s doubling trend line, SRAM scaling has slowed down in recent nodes and has fallen behind. In fact Intel’s own 10nm paper also demonstrates that.

SRAM Scaling Trend (Intel, IEDM 2017)

At the ISSCC 2018, Intel further detailed their SRAM array architecture.

128Kb SRAM Array Architecture (Intel, ISSCC 2018)

For their 10nm, Intel has significantly improved periphery area scaling resulting in double or better density improvement while having very little fin depopulation.

Periphery area scaling and fin population comparison (Intel, ISSCC 2018)
Test Shuttle

Intel’s 10nm test shuttle chip consisted of over 175 Mib of SRAM with at least 72 Mib of the low voltage cells and another 54 Mib of the high-density cells.


Die shot of Intel’s 10nm shuttle. (ISSCC 2018)

All in all, Intel has reported a bit density of 23.6 Mib/mm² and 20.4 Mib/mm² for the high-density and low-voltage cells respectively with around 78% array efficiency.

(Intel, ISSCC 2018)

Spotted an error? Help us fix it! Simply select the problematic text and press Ctrl+Enter to notify us.

Notify of
Newest Most Voted
Inline Feedbacks
View all comments
Maynard Handley
Maynard Handley
3 years ago

Nice article as usual, David, But might I suggest that in most places you replace “resistivity” with “resistance”?

The issue is not “resistivity” per se, no?
The issue is that what matters is the resistance of “wires as manufactured”. This resistance is a composite of bulk resistivity, surface effects, and the effective area that can be dedicated to the wire given manufacturing realities. Use of cobalt lowers the resistance because even though the bulk resistivity goes up, the other factors in that composite go down — the surface effects and the reduced effective area.

James L
James L
Reply to  Maynard Handley
3 years ago

“The issue is not “resistivity” per se, no?”

The issue is both resistance and resistivity though. Resistivity is an intrinsic property which is meant to describe the natural resistance to the flow in unconstrained space. But as you scale, you are no longer in unconstrained space and the increased surface scattering is said to affect the resistivity of the material.

I can’t speak for the author but as someone who discussed the subject with engineers in the past, often when they refer to the resistivity of the wire, it is a simple way of referring to the resistivty of the composite wire (core/barrier as a single material) rather than the resistance which is a function of the length and cross-sectional area.

Sanne Deijkers
Sanne Deijkers
2 months ago

Hi, very nice article which really give insight in the pros and cons of copper versus cobalt. I was however wondering where the following claim comes from: “Additionally, in contrast to copper, it has been demonstrated that a single film, as thin as 1 nm, is sufficient to serve as both the liner and barrier for cobalt.”

I would be very interested to learn more about the difference in barrier requirements for the use of copper and cobalt.

Would love your thoughts, please comment.x

Spelling error report

The following text will be sent to our editors: