deepseek-v4.1IMAGE
DGX Spark vs RTX 5090: local DeepSeek V4.1 Flash quant runs stable but slow
The author compares a quantized DeepSeek V4.1 Flash on DGX Spark with qwen3.8-flash-next via Strata on an RTX 5090: both run stably but noticeably slowly, forcing a capability-versus-speed trade-off.
PoTeto
@PoTeto_UGCCCArchived 5 days agoOriginally posted 2026.10.25
PoTeto @PoTeto_UGCCCOriginally posted 2026.10.25
ローカルLLM速度比較： DGX SparkのDeepSeek V4.1 flashの方は量子化版。RTX5090のqwen3.8-flash-nextはStrata。安定稼働はしているけれど、だいぶ遅い。。。。性能をとるかスピードをとるかは悩ましいところ。
x.com/poteto_ugccc/status/2114145245097247074Archived 2026.10.26 12:05