Akeana 5000 Series Processors
The Akeana 5000 Series of processors is highly configurable and customizable, supporting applications such as mobile computation, data center, and cloud networking. The Akeana 5000 Series is optimized to give industry-leading performance, is capable of supporting rich-OS (e.g. Android, Linux) application computation, and has the ability to scale to hundreds of coherent processors.
These processors can be configured with support for 1-, 2-, or 4-way Multi-threading.
The Akeana 5000 Series represents the high range of Akeana’s three processor product lines , offering customers one of the broadest processor IP portfolios. Start your journey today with one or more of these Akeana processors.
12-stage pipeline, Out-of-Order architecture
MMU with
2048-entry,
8-way TLB
Multi-threaded architecture support
6-wide to 10-wide issue architecture supported
Akeana 5000 Series Processor Diagram
Akeana 5000 Series Standard Configurations
Basic 5000 Series Configuration | Product | Standard Features | Typical Applications |
---|---|---|---|
RV64GCVBK_Zicbo + USH instruction set Full RVA23 RISC-V Profile Single and Double-Precision Floating-Point User Mode Supervisor Mode Hypervisor extension Vector extension (128 bits) Vector Crypto extension 12-stage, out-of-order pipeline 48 bits Virtual Address range 256K L2 cache Scalable to fully-coherent many-core clusters ECC support AXI/ACE (512 bits) Physical Memory Protection (PMP) with 16 entries MMU | Akeana 5100 | 4-way instruction dispatch L1 I-cache: 32 KB/core L1 D-cache: 32KB/core 33 bits Physical Address Space MMU with 512-entry, 4-way TLB | "Big" core in Big/Little configurations |
Akeana 5200 | 6-way instruction dispatch L1 I-cache: 32 KB/core L1 D-cache: 32 KB/core L2 cache prefetcher 39 bits Physical Address Space MMU with 1024-entry, 4-way TLB | “Big” core in Big/Little configurations |
|
Akeana 5300 | 8-way instruction dispatch L1 I-cache: 64 KB/core L1 D-cache: 64 KB/core L2 cache prefetcher 39 bits Physical Address Space MMU with 2048-entry, 8-way TLB | Datacenter / Infrastructure compute core |
Add-On Options Include
- L1 Instruction cache, up to 128 KB
- L1 Data cache, up to 64 KB
- L2 cache, up to 1 MB
- Physical Address width 32 to 56 bits
- Virtual Address width 39, 48, or 57 bits
- Shared, unified Last-Level Cache (LLC), up to 16 MB
- Scalar Cryptographic (Zk) extension
- Hypervisor (H) extension
- Vector (V) extension (64 to 512 bits)
- Vector Cryptographic (Zvk) extension
- Physical Memory Protection (PMP), up to 64 regions
- TLB up to 2048 entries, 8 ways
- Custom instructions
- Support for 2-way or 4-way multithreaded microarchitecture