Notebookcheck Logo

RISC-V specifications update brings improved machine learning, virtualization and encryption instructions

RISC-V is aiming for increased adoption (Image Source: RISC-V International)
RISC-V is aiming for increased adoption (Image Source: RISC-V International)
RISC-V International is pushing for increased adoption with the introduction of 15 new specifications for the RISC-V ISA. Most prominent are the Vector specs that expand support for AI / machine learning, the Scalar Cryptography specs that make it easy to add improved security features even on the smallest IoT devices, and the Hypervisor specs that add compatibility with other OSes for data center environments and allow for increased malware protection.

RISC-V International, the organization behind the standardization of the RISC-V architecture, announced yesterday the ratification of 15 new specifications, of which the most important are Vector, Scalar Cryptography and Hypervisor. These new specifications are meant to accelerate the adoption of the RISC-V ISA and provide improved features for emerging industries such as AI / machine learning, IoT, connected / autonomous cars and aircraft, data centers and beyond.

The Vector specs represent a set of just over 100 instructions that can help accelerate ML inference for audio vision, voice processing and other related AI applications implemented with the onboard computers on autonomous vehicles.

With the introduction of the Scalar Cryptography specs, adding cryptographic hash and block cipher algorithms to applications is now an order of magnitude faster than using standard instructions. This should help app developers to cheaply implement advanced encryption functions by default, even in the smallest IoT devices.

Improved compatibility with popular OSes and increased malware protection is also provided through the Hypervisor specs. RISC-V processors can now efficiently host type-1 (bare-metal) and type-2 (hosted) hypervisors, effectively driving increased adoption for the RISC-V architecture in cloud and embedded applications where virtualization is critical, such as data centers, automotive applications, and industrial control applications.

 

Buy the Sipeed Maixduino MCU with ESP32 Module for RISC-V AI + IoT on Amazon

Read all 2 comments / answer
static version load dynamic
Loading Comments
Comment on this article
Please share our article, every link counts!
.170
> Notebook / Laptop Reviews and News > News > News Archive > Newsarchive 2021 12 > RISC-V specifications update brings improved machine learning, virtualization and encryption instructions
Bogdan Solca, 2021-12- 3 (Update: 2021-12- 3)