New AMBA 5 ACE/AXI Specification: More About Atomic Transactions
February 23, 2018
As discussed in the previous installment of this blog, a new class of atomic transactions was introduced in the AMBA® 5 ACE/AXI specification to make operations at the remote locations more streamlined and efficient. We have considered an example of AtomicStore transaction with ADD operation and discussed why it was more efficient than relying on the older semaphore-like exclusive operations. In this installment of the blog we will take a broader look at all atomic transactions and review the changes in signaling.
To begin with, there are four (4) atomic transactions:
- AtomicStore
- AtomicLoad
- AtomicSwap
- AtomicCompare
Related Semiconductor IP
- CXL 3.0 Premium Controller EP/RP/DM/SW 128-1024 bits with AMBA bridge and Advanced HPC Features (Arm CCA)
- CXL 3.0 Premium Controller EP/RP/DM 1024b/512b/256b/128b with AMBA bridge for CXL.io and LTI & MSI Interfaces
- CXL 3.0 Premium Controller EP/RP/DM 1024b/512b/256b/128b with AMBA bridge for CXL.io
- CXL 2.0 Premium Controller Device/Host/DM 512b with AMBA bridge and Advanced HPC Features (Arm CCA)
- CXL 2.0 Premium Controller Device/Host/DM 512b with AMBA bridge + LTI and MSI-GIC interfaces
Related Blogs
- New AMBA 5 ACE/AXI Specification: Rationale for Atomic Transactions
- New AMBA 5 ACE/AXI Specification and Its Support in Cadence ACE/AXI VIP
- Why do I need an AMBA 5 CHI Memory Controller?
- Debug of AMBA AXI Outstanding Transactions
Latest Blogs
- Why Choose Hard IP for Embedded FPGA in Aerospace and Defense Applications
- Migrating the CPU IP Development from MIPS to RISC-V Instruction Set Architecture
- Quintauris: Accelerating RISC-V Innovation for next-gen Hardware
- Say Goodbye to Limits and Hello to Freedom of Scalability in the MIPS P8700
- Why is Hard IP a Better Solution for Embedded FPGA (eFPGA) Technology?