|
Qwen3.8-Flash-Next
|
|
317
|
37189
|
October 6, 2026
|
|
DeepSeek v4.1 Flash
|
|
220
|
31215
|
October 6, 2026
|
|
New Deepseek 4.1 flash recipe
|
|
10
|
1357
|
October 6, 2026
|
|
September 2026 Availability of GB10's
|
|
181
|
11964
|
October 6, 2026
|
|
Windows App (RDP on MacOS) for access to Spark ubuntu desktop - broken again
|
|
17
|
321
|
October 6, 2026
|
|
Sfn slot doesn't match unpacked one! PHY 0 L2->L1 0.0 vs L1->L2 0.3
|
|
18
|
147
|
October 6, 2026
|
|
cuDSS 0.8.0 uniform batch (UBATCH_SIZE > ~160): NaN solutions depending on the content of unrelated device memory
|
|
4
|
60
|
October 6, 2026
|
|
Use laya-mlx + laya for dgx spark to launch Snake Test on ASUS GX10
|
|
1
|
173
|
October 6, 2026
|
|
[bug] [nvcc] cicc dies with SIGILL lowering __nvvm_atom_add_gen_d for sm_120
|
|
0
|
12
|
October 6, 2026
|
|
Sharing an Open-Source (CC0) Architecture for Chip-Level MicroChannel Liquid Cooling (MicroChannel-DirectFlow) & Exploring Collaboration
|
|
0
|
13
|
October 6, 2026
|
|
Configuration files for Benetel O-RU
|
|
1
|
25
|
October 6, 2026
|
|
Swift 1.5 Qwen3.8-Flash-Next : 63.4% fewer thinking tokens
|
|
47
|
2391
|
October 6, 2026
|
|
Still waiting on our $10K AWS credits application (Ingest LLC)
|
|
1
|
40
|
October 5, 2026
|
|
A correctly rounded implementation of the half-precision arctangent function hatan()
|
|
0
|
24
|
October 5, 2026
|
|
Interval property not working on sgie models or models that use nvpreprocess on deepstream 9.1
|
|
0
|
15
|
October 5, 2026
|
|
DGX Spark RMA
|
|
3
|
475
|
October 5, 2026
|
|
Prioritizing Compute on the GB10
|
|
9
|
230
|
October 5, 2026
|
|
DGPP: a GB10-optimized C++/CUDA inference engine
|
|
100
|
5557
|
October 5, 2026
|
|
Aleph Alpha Kolibri-1 - interesting new model for german and english texts
|
|
8
|
480
|
October 5, 2026
|
|
Weird Post Turn System Prompt Injection Issues - Qwen 3.8 Flash
|
|
13
|
623
|
October 5, 2026
|
|
Qwen3-Next-80B on DGX Spark: 77 tok/s and 12 s load on one Spark, 121 tok/s and 10 s load on two
|
|
2
|
212
|
October 5, 2026
|
|
DeepSeek-V4-Flash-Vision-Exp on one DGX Spark: text + images, 245k context, CUDA graphs — plus a pack that keeps 90 % of the original model
|
|
16
|
1647
|
October 5, 2026
|
|
[Community project] Thinkube: Kubernetes done right on your own Spark
|
|
0
|
84
|
October 5, 2026
|
|
The Art of Multi-Node vLLM on DGX Spark: Recipes, Benchmarks, and Black Magic
|
|
7
|
238
|
October 5, 2026
|
|
JEV and friends
|
|
20
|
1455
|
October 5, 2026
|
|
2× DGX Spark DeepSeek-V4-Flash-0731 Stable Build
|
|
3
|
262
|
October 5, 2026
|
|
Single DGX-Spark - Qwen 3.8-Flash-Next at ~43tok/sec in Coding
|
|
79
|
15412
|
October 5, 2026
|
|
Dgx spark b10
|
|
0
|
143
|
October 5, 2026
|
|
Stable Qwen3.8-27B NVFP4 + DFlash2 on a Single DGX Spark
|
|
2
|
247
|
October 5, 2026
|
|
Student request for Brev credits - Isaac Sim robotics project
|
|
0
|
21
|
October 4, 2026
|