A chiplet[1][2][3][4] is a tiny integrated circuit (IC) that contains a well-defined subset of functionality. It is designed to be combined with other chiplets on an interposer in a single package to create a complex component such as a computer processor. Each chiplet in a computer processor provides only a portion of the processor's total functionality. A set of chiplets can be implemented in a mix-and-match "Lego-like" assembly. This contrasts with the traditional, monolithic system-on-a-chip (SoC) architecture, which integrates all functionality onto a single silicon die and provides several advantages over a traditional system on chip (SoC) which is monolithic as it comprises a single silicon die:
- Reusable IP (intellectual property):[5] the same chiplet can be used in many different devices
- Heterogeneous integration:[6] chiplets can be fabricated with different processes, materials, and nodes, each optimized for its particular function
- Known good die:[7] chiplets can be tested before assembly, improving the yield of the final device.
Multiple chiplets working together in a single integrated circuit may be called a multi-chip module, hybrid IC, 2.5D IC, or an advanced package.
Chiplets may be connected with standards such as UCIe, bunch of wires (BoW), AIB, OpenHBI, and OIF XSR.[8][9] Chiplets not designed by the same company must be designed with interoperability in mind.[10]
The term was coined by University of California, Berkeley professor John Wawrzynek as a component of the RAMP Project (research accelerator for multiple processors) in 2006 [11][12] extension for the Department of Energy.
Common examples include:
- Intel Meteor Lake
- AMD Ryzen based on Zen 2 and later architecture
- NVidia H100
See also
editReferences
edit- ^ Brookes (25 July 2021). "What Is a Chiplet?". How-To Geek. Retrieved 28 December 2021.
- ^ "Chiplet". WikiChip. Retrieved 28 December 2021.
- ^ AnySilicon "The Ultimate Guide to Chiplets" Retrieved 8 December 2024
- ^ Don Scansen, EE Times "Chiplets: A Short History Retrieved 5 December 2022
- ^ Keeler. "Common Heterogeneous Integration and IP Reuse Strategies (CHIPS)". DARPA. Retrieved 28 December 2021.
- ^ Kenyon (6 April 2021). "Heterogeneous Integration and the Evolution of IC Packaging". EE Times Europe. Retrieved 28 December 2021.
- ^ Bertin, Claude L.; Su, Lo-Soun; Van Horn, Jody (2001). "Known Good die (KGD)". Area Array Interconnection Handbook. SpringerLink. pp. 149–200. doi:10.1007/978-1-4615-1389-6_4. ISBN 978-1-4613-5529-8. Retrieved 7 October 2022.
- ^ "Waiting for Chiplet Standards". 25 March 2021.
- ^ "Is UCIe Really Universal?". 22 November 2022.
- ^ "UCIe Goes Back to the Drawing Board". 22 February 2024.
- ^ Patterson, D.A. (March 2006). "RAMP: Research accelerator for multiple processors - a community vision for a shared experimental parallel HW/SW platform". 2006 IEEE International Symposium on Performance Analysis of Systems and Software. pp. 1–. doi:10.1109/ISPASS.2006.1620784. ISBN 1-4244-0186-0.
- ^ Wawrzynek, John (2015-05-01). "Accelerating Science Driven System Design With RAMP". UCB. doi:10.2172/1186854. OSTI 1186854.
Further reading
edit- Clark, Don (11 May 2023). "U.S. Focuses on Invigorating 'Chiplets' to Stay Cutting-Edge in Tech". The New York Times.