RISC-V

1190 readers
2 users here now

RISC-V (pronounced "risk-five") is a license-free, modular, extensible instruction set architecture (ISA).

riscv.org

Youtube | Twitter

Matrix space

Other RISC-V communities on Lemmy

founded 5 years ago
MODERATORS
1
2
3
4
5
6
7
8
 
 

RVA23 profile of RISC-V marks a turning point in how mainstream CPUs are expected to scale performance. By making the RISC-V Vector Extension (RVV) mandatory, it elevates structured, explicit parallelism to the same architectural status as scalar execution. Vectors are no longer optional accelerators bolted onto speculation-heavy cores. They are baseline capabilities that software can rely on.

RVA23 doesn’t force scalar execution to become deterministic. It simply makes determinism viable because the scalar side is no longer responsible for throughput. The vector unit handles the parallel work explicitly, and the scalar core becomes a coordinator that can be simple, predictable, and low‑power without sacrificing performance.

To understand why this shift matters, it helps to recall how thoroughly speculative execution came to dominate high-performance CPU design. It delivered speed, but at increasing cost—in power, complexity, verification burden, and security exposure. RVA23 does not reject speculation. Instead, it restores balance. It acknowledges that predictable, vector-driven parallelism is now a credible, mainstream path for performance growth.

9
10
11
 
 

SpacemiT Key Stone K3 series chips adopt RISC-V isomorphic fusion computing technology, integrating SpacemiT's self-developed 8 high-performance computing cores X100 and 8 ultra-wide parallel computing AI cores A100, providing 130 KDMIPS general computing power and 60TOPS general AI computing power, and can run 30 billion parameter large models smoothly.

K3 series chips are mainly used in AI computers, intelligent robots, personal intelligent computing servers, edge intelligent computing servers, large model inference machines and other fields.

12
13
 
 

SpacemiT K3 is an upcoming RVA23-compliant 64-bit RISC-V processor based on X100 cores clocked at up to 2.5 GHz. So far, we had limited information, but SpacemiT gave remote access to one SpacemiT K3-powered server to Sander, and he was kind enough to share some system information and early benchmarks.

14
15
 
 

Xous is a message-passing microkernel implemented in pure Rust, targeting secure embedded applications. This talk covers three novel aspects of the OS: hardware MMU support (and why we had to make our own chip to get this feature), how and why we implemented the Rust standard library in Rust (instead of calling the C standard library, like most other Rust platforms), and how we combine the power of Rust semantics with virtual memory to create safe yet efficient asynchronous messaging primitives. We conclude with a short demo of the OS running on a new chip, the "Baochip-1x", which is an affordable, mostly-open RTL SoC built in 22nm TSMC, configured expressly for running Xous.

The world is full of small, Internet-of-Things (IoT) gadgets running embedded operating systems. These devices generally fall into two categories: larger devices running a full operating system using an MMU which generally means Linux, or smaller devices running without an MMU using operating systems like Zephyr, chibios, or rt-thread, or run with no operating system at all. The software that underpins these projects is written in C with coarse hardware memory protection at best. As a result, these embedded OSes lack the security guarantees and/or ergonomics offered by modern languages and best practices.

The Xous microkernel borrows concepts from heavier operating systems to modernize the embedded space. The open source OS is written in pure Rust with minimal dependencies and an emphasis on modularity and simplicity, such that a technically-savvy individual can audit the code base in a reasonable period of time. This talk covers three novel aspects of the OS: its incorporation of hardware memory virtualization, its pure-Rust standard library, and its message passing architecture.

Desktop OSes such as Linux require a hardware MMU to virtualize memory. We explain how ARM has tricked us into accepting that MMUs are hardware-intensive features only to be found on more expensive “application” CPUs, thus creating a vicious cycle where cheaper devices are forced to be less safe. Thanks to the open nature of RISC-V, we are able to break ARM’s yoke and incorporate well-established MMU-based memory protection into embedded hardware, giving us security-first features such as process isolation and encrypted swap memory. In order to make Xous on real hardware more accessible, we introduce the Baochip-1x, an affordable, mostly-open RTL 22nm SoC configured expressly for the purpose of running Xous. The Baochip-1x features a Vexriscv CPU running at 400MHz, 2MiB of SRAM, 4MiB of nonvolatile RRAM, and a quad-core RV32E-derivative I/O accelerator called the “BIO”, based on the PicoRV clocked at 800MHz.

Most Rust targets delegate crucial tasks such as memory allocation, networking, and threading to the underlying operating system’s C standard library. We want strong memory safety guarantees all the way down to the memory allocator and task scheduler, so for Xous we implemented our standard library in pure Rust. Adhering to pure Rust also makes cross-compilation and cross-platform development a breeze, since there are no special compiler or linker concerns. We will show you how to raise the standard for “Pure Rust” by implementing a custom libstd.

Xous combines the power of page-based virtual memory and Rust’s strong borrow-checker semantics to create a safe and efficient method for asynchronous message passing between processes. This inter-process communication model allows for easy separation of different tasks while keeping the core kernel small. This process maps well onto the Rust "Borrow / Mutable Borrow / Move" concept and treats object passing as an IPC primitive. We will demonstrate how this works natively and give examples of how to map common programming algorithms to shuttle data safely between processes, as well as give examples of how we implement features such as scheduling and synchronization primitive entirely in user space.

We conclude with a short demo of Xous running on the Baochip-1x, bringing Xous from the realm of emulation and FPGAs into everyday-user accessible physical silicon.

16
17
18
19
20
 
 

Box64 is a x86 emulator that supports RiSC-V. With this, its possible to run steam, wine/proton, many games, and a lot of software on a RiSC-V computer!

21
22
23
24
 
 

RISC-V is an industry standard, like USB or Wi-Fi. The specifications are publicly available under the Creative Commons license and every engineer, wherever they are in the world, can use them to design their products locally, while engaging with the global RISC-V ecosystem.

This standard is defined by RISC-V International and its members. Decisions are voted upon collectively, ensuring every member is heard. It’s a model that has worked for us for many years, ensuring any updates to the RISC-V ISA happen transparently, without breaking existing designs, and always in service of the broader ecosystem.

The RISC-V ISA is already an industry standard and the next step is impartial recognition from a trusted international organization.

Today, I’m excited to announce that we have taken that first step. RISC-V International has been approved as a recognized PAS (that’s publicly available specification) Submitter by the ISO/IEC Joint Technical Committee (JTC 1).

This means we’re able to submit draft international papers, starting with the The RISC-V Instruction Set Manual, for consideration as true, international standards.

25
 
 

One of the more unexpected talks at last week's Ubuntu Summit 25.10 in London was by Antonio Salvemini of Bolt Graphics, who introduced the company's forthcoming range of Zeus graphics accelerator hardware. These are very unlike any conventional GPUs – or indeed anything else.

[…]

Bolt's Zeus hardware will use an entirely different model, and we found it refreshing that the company's How it works page doesn't mention the dreaded initialism "AI" once. These accelerators are aimed at producing graphics using a specific rendering method called path tracing.

[…] Path tracing is a step further on from simple ray tracing. A few decades ago, ray tracing was a favorite way to demonstrate high-resolution, multi-color computer graphics, taking hours to days to render scenes of shiny spheres.

As Salvemini put it: "The problem with ray tracing is that each light wave only bounces one way. In path tracing, they can bounce anywhere, and you randomly select just some of these paths to display." Thus, Monte Carlo path tracing (MCPT) – as described in this 2024 UCSD computer graphics lecture [PDF] – uses Monte Carlo simulation, as invented by John von Neumann and Stanislaw Ulam during World War II.

view more: next ›