Flash MX NVMe expansion
The Flash MX IP Core implements a hardware accelerator for zstd compression and decompression. Compute-intensive software-based compression is offloaded from the host to the IP, delivering high compression performance at unmatched power efficiency.
Overview
The Flash MX IP Core implements a hardware accelerator for zstd compression and decompression. Compute-intensive software-based compression is offloaded from the host to the IP, delivering high compression performance at unmatched power efficiency.
Standards
Hardware accelerated zstd
Compression algorithm: zstd
Interface: AXI4, CHI
Architecture
Modular architecture, enables scalability to meet customer throughput requirements
Architectural configuration parameters accessible to fine tune performance
HDL Source Licenses
Synthesizable System Verilog RTL (encrypted)
Implementation constraints
UVM testbench (self-checking)
Vectors for testbench and expected results
User Documentation
Features
Turn key solution: compression, compaction, memory management
Transparent addressing to operating system and applications
Operates on page granularity to enable high compression performance
Deliverables
Performance evaluation license
C++ compression model for integration in customer performance simulation model
FPGA evaluation license
Encrypted IP delivery (Xilinx)
Applications
Today costly host CPU cyscles are allocated to software based compression, when data is compressed before sent off over the interconnect to storage. The product benefit is more system performance at less power when the host is off-loaded with the software based compression, at the same time as the storage interconnect make up to 50% more bandwidth available and the storage device store 2-4x the amount of data.
Integration
Flash MX is integrated on the SoC as a hardware accelerator, a master node on the SoC interconnect. This could be either on the Host CPU or on the DPU / SmartNIC. Part of the integration includes a software driver.
Benefits
High performance and low latency hardware accelerated zstd at unmatched power efficiency. Off-loading CPU – More cycles to released to user work loads. Power efficiency – Less energy. Speed – Fast compression and low latency access. Compatible with AXI4/CHI, both 128-b and 256-b bus interface.
Performance / KPI
Feature | Performance |
Compression ratio: | 2-4x across diverse data sets |
Compression throughput: | (contact for details) |
Decompression throughput: | (contact for details) |
Frequency: | 1.2GHz (@TSMC 5nm) |
IP area: | (Pending development details) |
Cache MX
The Cache MX compression solution increases the cache capacity by 2x at an 80% area and power saving to comparable SRAM capacity.
SuperRAM
High performance and low latency hardware accelerated compression at unmatched power efficiency.
Ziptilion™ BW
Delivers up to 25% more (LP)DDR bandwidth at nominal frequency and power, enabling a significantly more performance and energy efficient SoC.
DenseMem
Double the CXL connected memory capacity with data DenseMem.
Flash MX
Double the NvMe storage capacity with Flash MX.
SphinX
High Performance and Low Latency industry-standard encryption / decryption. Independent non-blocking encryption and decryption channels.