IP for High-Bandwidth Applications

High-Performance, High-Bandwidth IP Platforms & 2.5D Solutions for Networking, High-Performance Computing and AI/ML

Cloud, data center, networking and AI applications are resulting in an explosion of data being generated, moved, stored and analyzed. Building ASICs that support the massive need for memory and processing requires working in advanced FinFET-class solutions such as high-bandwidth memory (HBM2) stacks, 2.5D packaging and IP optimized specifically for high-performance, high-bandwidth markets. It requires doing things in silicon that have never been done before.

eSilicon specializes in high-performance, high-bandwidth IP + 2.5D solutions that target networking, high-performance computing and artificial intelligence (AI) applications. eSilicon’s specialized IP ecosystem in this category includes a high-performance, extremely configurable 56G and 112G SerDes family; a ternary content-addressable (TCAM) compiler; a configurable, high-bandwidth memory (HBM2) PHY, multiple network-optimized memories; and extended-voltage general-purpose and LVDS I/Os.

Highly Configurable 7nm Networking IP platform

eSilicon’s 7nm IP platform delivers a complete ecosystem of networking-optimized IP with high configurability designed in. All IP in the platform is “plug and play,” using the same metal stack, reliability requirements, operating ranges, control interfaces and DFT methodology. This configurability and compatibility results in better performance, higher density and faster time to market.

This is our second-generation platform with architectural enhancements from our silicon-proven platform in previous FinFET technologies. It was designed specifically to meet the high-speed, high-bandwidth requirements of ASICs and ASSPs targeted for networking and high-performance computing applications.

Additional features and benefits for eSilicon’s 7nm IP platform are available under NDA. Contact sales@esilicon.com. Front-end views of our 7nm IP offering are available now by contacting sales@esilicon.com.

14/16nm IP

Our silicon-proven 14/16nm specialized memories provide 2.5GHz worst-case operation with more than a billion searches per second along with the 2.5D integration of 1024 Gbytes/sec data rate high-bandwidth memory (HBM2).

56G/112G 7nm SerDes

At the core of our 7nm IP platform is eSilicon’s SerDes technology, a new breed of performance and versatility based on a novel DSP-based architecture. Two 7nm PHYs support 56G and 112G NRZ/PAM4 operation to provide the best power efficiency tradeoffs for server, fabric and line-card applications. The architecture delivers unprecedented power efficiency for a true long-reach capability, with hole-free operation down to 1 Gb/s. The clocking architecture provides extreme flexibility to support multi-link and multi-rate operations per SerDes lane. A multitude of protocols are supported including Ethernet and Fibre Channel. The architecture allows scaling power consumption even further for shorter-reach channels.

High-Bandwidth Memory

HBM2 achieves higher bandwidth while consuming less power in a substantially smaller form factor than DDR4, or GDDR5. At 14/16nm, HBM2 addresses the bandwidth gap with up to 256 GB/s data rate per memory at 2Gbps pin speed. At 7nm, the pin speed increases to 2.4Gbps with bandwidth up to 307GB/s.  The HBM2 PHY interface features eight independent channels using a total of 1024 data pins. With support for 2, 4, or 8 HBM2 stacks, the density of signals, coupled with interposer design, requires careful design, thorough timing analysis and validation. eSilicon’s HBM2 PHY is a complete, validated hardened IP that is ready for chip integration.

eSilicon’s 2.5D/HBM2 is first to production with a comprehensive 2.5D solution that includes eSilicon’s HBM2 PHY, interposer  and systems in package (SiP) design; volume manufacturing, assembly and test; and complete 2.5D/HBM ecosystem management.

Why TCAM?

Memory processing tends to be the bottleneck in network performance. If memory cannot keep up with increasingly fast processors, the processors have to wait, stalling the system. Specialty memories address the problem in a variety of ways. Ternary content-addressable memories (TCAMs) are unique ─ they search an entire lookup table in one cycle. eSilicon’s 7nm ternary CAM compiler provides more than a 50% increase in performance compared to previous FinFET generations, enabling high-efficiency, cost-effective solutions for applications such as network search engines, cache for network processors, QoS services, classifications, Ethernet, ATM switches and other diverse networking applications.

Fast Cache

eSilicon has designed custom high-speed single-port cache memories that are optimized to meet the high-performance requirements of industry-standard processor cores. Our 14LPP high-speed single-port fast cache compiler is part of our complete 14LPP IP platform. eSilicon’s new 7nm fast cache compiler offers  more than a 35% increase in performance from previous FinFET generations.

Multi-Port Register File & Asynchronous Register File

eSilicon has designed four-port register file memory compilers and asynchronous register file memory compilers catering to multiple customers in the networking and communications domains. The four-port memories have two write ports and two read ports. The asynchronous memories have one synchronous write port and one asynchronous read port.

Networking IP Platforms

14LPP Products

High-speed SP TCAM
Ultra-high-speed SP TCAM
High-speed SP fast cache
High-speed DP SRAM
High-speed pseudo 2-port SRAM
High-density pseudo 2-port SRAM
High-density 2-port asynchronous RF
Ultra-high-speed pseudo 2-port SRAM
High-speed four-port register file
1.8V oxide 1.8V/2.5V/3.3V general-purpose I/O library
1.8V oxide 1.8V LVDSOUT I/O library
1024-bit HBM2 PHY

7FF Products

56G and 112G long-reach SerDes
1024-bit HBM2 PHY
High-speed single-port ternary CAM (SP TCAM)
High-speed single-port fast cache (FC)
High-speed single-port (SP) SRAM
High-speed dual-port (DP) SRAM
High-speed 2-port asynchronous register file (2PARF)
High-speed pseudo 2-port (P2P) SRAM
Ultra-high-density (UHD) pseudo 2-port (P2P) SRAM
High-speed pseudo 4-port (P4P) SRAM
High-speed pseudo quad-port (PQP) SRAM
1.8V oxide 1.8V LVDS I/O library
1.8V oxide 1.8V/2.5V/3.3V general-purpose I/O library

Detailed Information

Front-end views, silicon quality results, white papers or complete data sheets, are available from sales@esilicon.com. Additional features and benefits for eSilicon’s 7nm IP platform are available under NDA. Contact sales@esilicon.com.

Online Access

All eSilicon IP is available in IP Navigator at https://star.esilicon.com, our online IP exploration tool. Navigator provides access to eSilicon’s full portfolio of IP products. Memory instances may be generated, analyzed and downloaded. Power, performance, and area (PPA) data is pre-loaded for easy data comparison and analysis.

Resources

White Papers

  • The New Deep Learning Memory Architectures You Should Know About | eSilicon
  • Supercharge Your Applications with Samsung High-Bandwidth Memory | Samsung Memory
  • Multi-Die Packaging and Thermal Superposition Modeling | Amkor, eSilicon & Samsung Foundry for Semitherm 2018
  • Start Your HBM/2.5D Design Today | eSilicon, SK Hynix, Amkor Technology, Northwest Logic and Avery Design Systems

Videos & Webinar Replays

Brochures:

  • 7FF IP platform for networking, high-performance computing, AI & 5G
  • 14LPP IP platform for networking, high-performance computing, AI & 5G
  • Ternary CAMs (TCAMs)
  • 28/16/14/7nm HBM2 PHY