random RISC-V assembly test generation

Concept

Random RISC-V assembly test generation is evidenced through RISCV-DV, a Google-developed tool used to generate random RISC-V assembly tests. In the cited verification environment, generated tests supplied vector instructions for a VPU, with adaptations for RVV 0.7.1, vector configuration instructions, memory constraints, data-page initialization, and incremental instruction enablement.

First seen 5/27/2026

Last seen 6/1/2026

Evidence 1 chunks

Wiki v1

WIKI

random RISC-V assembly test generation

Random RISC-V assembly test generation is represented in the evidence by the use of RISCV-DV, a Google-developed tool that generates random RISC-V assembly tests. In the cited verification environment, those generated tests were used to provide vector instructions to a VPU under test. [C1]

Role in vector verification

READ FULL ARTICLE →

NEIGHBORHOOD

No graph connections found for this entity yet. It may appear in future ingestion runs.

explore full graph →

RELATIONSHIPS

1 connections

riscv-dv ← implements 99% 1e

RISCV-DV is designed to generate random RISC-V assembly tests.

LINKED ENTITIES

2 links

riscv-dv IMPLEMENTS Extracted graph relationship

spike USED_WITH The evidence describes Spike being used in the same verification environment to execute scalar instructions, provide vector instructions to UVM, and act as a golden/reference model.

CITATIONS

5 sources

5 citations — click to expand

[1] RISCV-DV is a Google-developed tool that generates random RISC-V assembly tests and was used to provide vector instructions to a VPU. source

[2] RISCV-DV targeted a later RVV version than 0.7.1, so parts were developed and adapted to fit the target environment. source

[3] Reported RISCV-DV additions included vsetvli generation, memory-operation changes for element width and vector length, data-page initialization selection, constrained memory addresses, and RVV 0.7.1 adaptation. source

[4] Instructions were initially blacklisted from generated tests during development and gradually enabled as errors were fixed. source

[5] Spike executed scalar instructions, provided vector instructions to UVM in program order, and acted as a golden/reference model for checking DUT results. source