AI expands HBM footprint
By Gary Hilson, EETimes (January 20, 2022)
High bandwidth memory (HBM) is becoming more mainstream. With the latest iteration’s specifications approved, vendors in the ecosystem are gearing to make sure it can be implemented so customers can begin to design, test and deploy systems.
The massive growth and diversity in artificial intelligence (AI) means HBM is less than niche. It’s even become less expensive, but it’s still a premium memory and requires expertise to implement. As a memory interface for 3D-stacked DRAM, HBM achieves higher bandwidth while using less power in a form factor that’s significantly smaller than DDR4 or GDDR5 by stacking as many as eight DRAM dies with an optional base die which can include buffer circuitry and test logic.
Like all memory, HBM makes advances in performance improvement and power consumption with every iteration. A key change when moving to HBM3 from HBM2 will a 100% performance improvement in the data transfer rate from 3.2/3.6Gbps to 6.4Gbps max per pin, said Jinhyun Kim, principal engineer with Samsung Electronics’ memory product planning team.
A second fundamental change is a 50% increase in the maximum capacity from 16GB (8H) to 24GB (12H). Finally, HBM3 implements on-die error correction code as an industry-wide standard, which improves system reliability, Kim said. “This will be critical for the next generation of artificial intelligence and machine learning systems.”
E-mail This Article | Printer-Friendly Page |
Related News
- MIPS Expands Global Footprint with New Design Center and Talent for Systems Architects and AI Compute
- Tenstorrent Expands Deployment of Arteris' Network-on-Chip IP to Next-Generation of Chiplet-Based AI Solutions
- SensiML Expands Platform Support to Include the RISC-V Architecture
- Ceva Expands its Market Share Leadership in Wireless Connectivity IP, Strengthening its Solutions for Smart Edge AI/IoT applications
- Altair Signs Agreement to Acquire Metrics Design Automation Inc. Expands Footprint in EDA Industry
Breaking News
- Achronix and BigCat Wireless Collaborate to Deliver Unprecedented Power Efficiency and Performance for 5G/6G Wireless Applications
- Renesas Unveils Industry's First Automotive Multi-Domain SoC Built with 3-nm Process Technology
- LG and Tenstorrent Expand Partnership to Enhance AI Chip Capabilities
- CHERI Alliance Officially Launches, Adds Major Partners including Google, to Tackle Cybersecurity Threats at the Hardware Level
- Silicon Creations Celebrates Milestone with Delivery of 1,000th Production License for Fractional-N PLL
Most Popular
- Flex Logix Acquired By Analog Devices
- Arteris and MIPS Partner on High-Performance RISC-V SoCs for Automotive, Datacenter and Edge AI
- NanoXplore acquires Dolphin Design's ASIC business and strengthens its strategic position in aerospace
- SEMIFIVE Collaborates with Synopsys to Develop Advanced Chiplet Platform for High-Performance Multi-Die Designs
- Imagination DXS GPU officially certified as ASIL-B compliant