Statements (56)
| Predicate | Object |
|---|---|
| gptkbp:instance_of |
gptkb:software
|
| gptkbp:category |
linear algebra
|
| gptkbp:developed_by |
gptkb:DJ
|
| https://www.w3.org/2000/01/rdf-schema#label |
Roc SOLVER
|
| gptkbp:includes |
LU factorization
Cholesky factorization QR factorization |
| gptkbp:is_available_on |
open-source license
|
| gptkbp:is_compatible_with |
gptkb:Graphics_Processing_Unit
gptkb:software gptkb:Py_Torch gptkb:HIP |
| gptkbp:is_designed_for |
low latency
scientific computing high throughput |
| gptkbp:is_designed_to |
enhance scalability
improve accuracy reduce computational time support heterogeneous computing facilitate parallel processing maximize performance on AMD hardware support large datasets |
| gptkbp:is_integrated_with |
other RO Cm libraries
|
| gptkbp:is_maintained_by |
NVIDIA community
|
| gptkbp:is_optimized_for |
high-performance computing
|
| gptkbp:is_part_of |
NVIDIA's software stack
NVIDIA's AI software stack |
| gptkbp:is_used_in |
gptkb:academic_research
gptkb:Research_Institute industrial applications machine learning applications data science numerical simulations deep learning frameworks government research |
| gptkbp:is_utilized_in |
gptkb:technology
image processing data analytics financial modeling signal processing engineering simulations |
| gptkbp:language |
gptkb:C++
|
| gptkbp:offers |
performance optimization for AMDGP Us
|
| gptkbp:part_of |
Radeon Open Compute (RO Cm) ecosystem
|
| gptkbp:platform |
gptkb:operating_system
|
| gptkbp:provides |
API for developers
solvers for linear systems batched solvers |
| gptkbp:released |
gptkb:2019
|
| gptkbp:suitable_for |
real-time applications
|
| gptkbp:supports |
GPU acceleration
multi-GPU configurations multiple data types |
| gptkbp:uses |
CUDA technology
|
| gptkbp:bfsParent |
gptkb:HIP
|
| gptkbp:bfsLayer |
4
|