Skip to content
STIMSMITH

HARTBREAKER

Tool
First seen 7/3/2026
Last seen 7/3/2026
Evidence 18 chunks

NEIGHBORHOOD

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

explore full graph →

RELATIONSHIPS

21 connections
RTL Simulation uses → 100% 3e
HARTBREAKER uses RTL simulation to test CPU designs.
XiangShan evaluates → 100% 3e
HARTBREAKER was evaluated on the XiangShan RISC-V CPU.
litmus tests uses → 95% 3e
HARTBREAKER uses litmus tests to verify memory consistency model compliance.
Instruction Set Simulator uses → 100% 3e
HARTBREAKER uses an instruction set simulator (Spike) as a golden model.
NaxRiscv evaluates → 100% 3e
HARTBREAKER was evaluated on the NaxRiscv CPU.
Rocket evaluates → 100% 3e
HARTBREAKER was evaluated on the Rocket RISC-V CPU.
BOOM evaluates → 100% 3e
HARTBREAKER was evaluated on the BOOM RISC-V CPU.
Toooba evaluates → 100% 3e
HARTBREAKER was evaluated on the Toooba RISC-V CPU.
instruction throughput uses → 90% 2e
HARTBREAKER is designed for high instruction throughput.
HARTBREAKER paper ← introduces 100% 2e
The HARTBREAKER paper introduces the HARTBREAKER multi-hart RISC-V CPU fuzzer.
determinism anchors implements → 100% 2e
HARTBREAKER implements determinism anchors to manage non-determinism.
control-flow anchors implements → 100% 2e
HARTBREAKER implements control-flow anchors to handle control-flow non-determinism.
data-flow anchors implements → 100% 2e
HARTBREAKER implements data-flow anchors to handle data-flow non-determinism.
synchronization anchors implements → 100% 2e
HARTBREAKER implements synchronization anchors to bound non-determinism for verification.
Pre-silicon fuzzing implements → 95% 2e
HARTBREAKER is a pre-silicon fuzzer for multi-hart RISC-V CPUs.
spike uses → 100% 2e
HARTBREAKER uses Spike as a golden model ISS.
Cascade ← compares with 90% 2e
HARTBREAKER is compared to Cascade in terms of instruction diversity.
differential testing uses → 85% 1e
HARTBREAKER, like other fuzzers, relies on differential testing using an ISS or reference hart.
riscv-dv compares with → 100% 1e
HARTBREAKER is compared against RISCV-DV in terms of coverage and capabilities.
mret instruction mentions → 90% 1e
HARTBREAKER discovered a bug related to mret instruction handling in Toooba.
Verilator uses → 100% 1e
HARTBREAKER uses Verilator to simulate designs under test.