Abstract:We propose Sectored DRAM, a new, low-overhead DRAM substrate that reduces wasted energy by enabling fine-grained DRAM data transfers and DRAM row activation. Sectored DRAM leverages two key ideas to enable fine-grained data transfers and row activation at low chip area cost. First, a cache block transfer between main memory and the memory controller happens in a fixed number of clock cycles where only a small portion of the cache block (a word) is transferred in each cycle. Sectored DRAM augments the memory controller and the DRAM chip to execute cache block transfers in a variable number of clock cycles based on the workload access pattern with minor modifications to the memory controller's and the DRAM chip's circuitry. Second, a large DRAM row, by design, is already partitioned into smaller independent physically isolated regions. Sectored DRAM provides the memory controller with the ability to activate each such region based on the workload access pattern via small modifications to the DRAM chip's array access circuitry. Activating smaller regions of a large row relaxes DRAM power delivery constraints and allows the memory controller to schedule DRAM accesses faster. Compared to a system with coarse-grained DRAM, Sectored DRAM reduces the DRAM energy consumption of highly-memory-intensive workloads by up to (on average) 33% (20%) while improving their performance by up to (on average) 36% (17%). Sectored DRAM's DRAM energy savings, combined with its system performance improvement, allows system-wide energy savings of up to 23%. Sectored DRAM's DRAM chip area overhead is 1.7% the area of a modern DDR4 chip. We hope and believe that Sectored DRAM's ideas and results will help to enable more efficient and high-performance memory systems. To this end, we open source Sectored DRAM at <a class="link-external link-https" href="https://github.com/CMU-SAFARI/Sectored-DRAM" rel="external noopener nofollow">this https URL</a>.

What problem does this paper attempt to address?

### Problems Addressed by the Paper The paper aims to address the issue of energy waste in modern computer systems caused by coarse-grained memory access mechanisms. Specifically, the paper proposes a new architecture called **Sectored DRAM** to achieve fine-grained DRAM data transfer and row activation, thereby reducing energy consumption and improving performance. #### Core Issues 1. **Unnecessary data movement due to coarse-grained data transfer**: Standard DRAM interfaces transfer data at the cache block granularity, but most of the data in the cache block is not actually used, leading to energy waste. 2. **Coarse-grained row activation leading to the activation of many DRAM cells**: Modern memory-intensive workloads often have low spatial locality and cannot fully utilize the entire DRAM row, thus activating all cells wastes energy. #### Solution The paper proposes **Sectored DRAM**, a new low-overhead DRAM architecture that achieves fine-grained data transfer and row activation through the following two key technologies: 1. **Variable Burst Length (VBL)**: Allows dynamic adjustment of the data transfer length based on the actual needs of the workload. 2. **Sectored Activation (SA)**: Allows independent activation of small regions within a DRAM row (called "sectors"), thereby reducing energy consumption during activation. #### Main Contributions - **Performance and energy improvement**: By achieving fine-grained data transfer and row activation, Sectored DRAM can significantly reduce the energy consumption of memory-intensive workloads and improve system performance. - **Low hardware overhead**: The hardware overhead of Sectored DRAM is only 1.7% of the area of modern DDR4 chips and does not require modifications to the physical DRAM interface. - **Integration technology**: Developed two hardware technologies (Load/Store Queue Lookahead and Sector Predictor) to effectively integrate Sectored DRAM into existing systems, reducing the number of high-latency memory accesses. - **Open-source project**: The paper open-sources Sectored DRAM so that other researchers can reproduce the results and further study it. Through these improvements, Sectored DRAM demonstrates higher performance and lower energy consumption compared to traditional coarse-grained DRAM in multiple benchmark tests.

Sectored DRAM: A Practical Energy-Efficient and High-Performance Fine-Grained DRAM Architecture

Self-Managing DRAM: A Low-Cost Framework for Enabling Autonomous and Efficient in-DRAM Operations

Understanding the Security Benefits and Overheads of Emerging Industry Solutions to DRAM Read Disturbance

CR-DRAM: Improving DRAM Refresh Energy Efficiency with Inter-Subarray Charge Recycling

DR DRAM: Accelerating Memory-Read-Intensive Applications

Enabling Efficient and Scalable DRAM Read Disturbance Mitigation via New Experimental Insights into Modern DRAM Chips

CDAR-DRAM - an In-situ Charge Detection and Adaptive Data Restoration DRAM Architecture for Performance and Energy Efficiency Improvement.

Reducing Performance Impact of DRAM Refresh by Parallelizing Refreshes with Accesses

Improving DRAM Performance by Parallelizing Refreshes with Accesses

Half-DRAM: A High-Bandwidth and Low-Power DRAM Architecture from the Rethinking of Fine-Grained Activation

Retrospective: RAIDR: Retention-Aware Intelligent DRAM Refresh

SIMDRAM: a framework for bit-serial SIMD processing using DRAM

CODIC: A Low-Cost Substrate for Enabling Custom In-DRAM Functionalities and Optimizations

CDS-RSRAM: a Reconfigurable SRAM Architecture to Reduce Read Power with Column Data Segmentation

A Stacked Embedded DRAM Array for LPDDR4/4X Using Hybrid Bonding 3D Integration with 34gb/s/1gb 0.88pj/b Logic-to-Memory Interface

CREAM: A Concurrent-Refresh-Aware DRAM Memory architecture

Improving DRAM Performance, Reliability, and Security by Rigorously Understanding Intrinsic DRAM Operation

Architecture design exploration of three-dimensional (3D) integrated DRAM

Cryogenic quasi-static embedded DRAM for energy-efficient compute-in-memory applications

TDRAM: Tag-enhanced DRAM for Efficient Caching

FASA-DRAM: Reducing DRAM Latency with Destructive Activation and Delayed Restoration