|
cuBLASLt returns silently wrong FP16/BF16/FP8 GEMM results on Hopper under CUDA_MPS_ACTIVE_THREAD_PERCENTAGE (algoId 66, cluster kernels)
|
|
2
|
54
|
October 6, 2026
|
|
How can I verify whether a process is actually using my GPU/CUDA?
|
|
2
|
70
|
October 4, 2026
|
|
TensorFlow 2.21 GPU wheel for JetPack 7.2 / Jetson Orin
|
|
2
|
57
|
September 29, 2026
|
|
Dependency errors pulling down torch
|
|
14
|
337
|
September 22, 2026
|
|
TensorFlow 2.21 GPU wheel for JetPack 7.2 / Jetson Orin
|
|
0
|
32
|
September 22, 2026
|
|
Crash when trying to use Opencv Cuda functions on Jetson
|
|
6
|
95
|
September 21, 2026
|
|
RTX PRO 4000 Blackwell mixed-precision training: FP16/BF16, FP8, FP6, FP4 and FP32 accumulation
|
|
0
|
115
|
September 11, 2026
|
|
Under MPS, reclaiming a tenant wedged in the cuDNN sm100 SDPA decode kernel silently stalls other tenants — no error, no recovery
|
|
0
|
33
|
September 9, 2026
|
|
[WITHDRAWN] Qwen3.8-Flash-Next NVFP4 on 2x GB10: long-agent crash isolation and historical TP2 measurements
|
|
2
|
661
|
August 31, 2026
|
|
BF16 strongly-typed engine build segfaults in Myelin on RTX 5090 (sm_120) — identical network builds fine as FP16; reproduced on TensorRT 11.0 / 11.1
|
|
3
|
108
|
August 31, 2026
|
|
Docker image L4T JetPack 7.2.1 -- L4T r39.2.1
|
|
4
|
405
|
August 31, 2026
|
|
Repeated Xid 31 MMU faults during AMP-enabled YOLOv8s training on DGX Spark (GB10)
|
|
3
|
225
|
August 28, 2026
|
|
Aliased I/O TensorRT plugin bug
|
|
1
|
63
|
August 28, 2026
|
|
TensorRT 10.14 silently produces wrong detection scores for D-FINE (DETR-style) models on RTX 5090 / sm_120 — default build drops 12 detections to 0,
|
|
2
|
133
|
August 28, 2026
|
|
GB10 / DGX Spark: CUDNN_FE failure 11 and CUBLAS_STATUS_INTERNAL_ERROR under batch load - fixed by driver 580.173
|
|
1
|
226
|
August 26, 2026
|
|
Is Clang 18 unsupported for compiling with CUDA 12.6?
|
|
2
|
87
|
August 26, 2026
|
|
ONNX Runtime GPU inference on DGX Spark (GX10) — build guide and prebuilt binaries
|
|
1
|
1070
|
August 22, 2026
|
|
[Blackwell] cuDNN fused FP16 SDPA decode kernel (q_len=1) hangs under MPS multi-tenancy
|
|
0
|
43
|
August 21, 2026
|
|
GPU Utilization Bottleneck: Single-Process with 8 Streams vs. Multi-Process with 1 Stream each (DeepStream 7.1 / PyServiceMaker)
|
|
14
|
438
|
August 17, 2026
|
|
AOT tensorrt.plugin reloads plugin PTX via cuModuleLoadData on every enqueue in dynamic-shape engines
|
|
2
|
137
|
August 6, 2026
|
|
Techniques to Imporve TensorRT Model Inference Speed
|
|
3
|
312
|
August 5, 2026
|
|
TensorRT Engine Creation Fails with “tensor volume exceeds 2147483648” on MobileNet SSD v2 in DeepStream 7.1 (Jetson Orin Nano)
|
|
2
|
135
|
August 4, 2026
|
|
Quadro RTX 6000 does not handle BF16? Please make an update?
|
|
17
|
2771
|
August 3, 2026
|
|
INT8 Tensor Core corruption on Turing sm_75 cuBLAS llama.cpp BPE tokenizer
|
|
3
|
310
|
July 31, 2026
|
|
cuDNN AveragePool returns CUDNN_STATUS_NOT_SUPPORTED for tensors with batch dimension >= 65536 on RTX 5090
|
|
1
|
78
|
July 30, 2026
|
|
JetPack 5.1.7/L4T 35.6.5 is now live
|
|
1
|
236
|
July 30, 2026
|
|
JetPack 5.1.7/L4T 35.6.5 is now live
|
|
0
|
167
|
July 24, 2026
|
|
JetPack 5.1.7/L4T 35.6.5 is now live
|
|
0
|
198
|
July 24, 2026
|
|
JetPack 5.1.7/L4T 35.6.5 is now live
|
|
0
|
164
|
July 24, 2026
|
|
JetPack 5.1.7/L4T 35.6.5 is now live
|
|
0
|
169
|
July 24, 2026
|