// RUN โ€” FITS

Can the NVIDIA DGX Spark run AntiFormer?

Yes โ€” here's how fast and up to what context, with real measured numbers.

AntiFormer on the NVIDIA DGX Spark

QuantFits?Max contextSpeedBeyond context
FP16 yes 1734k 4039-4813.6 t/s ~147.9-176.3 (slow)
Q8 yes 1735k 8078-9627.3 t/s ~295.9-352.6 (slow)
Q4 yes 1735k 14425.1-17191.5 t/s ~528.4-629.7 (slow)

Estimate (memory-bound). Beyond "max context" the KV cache spills to SSD swap and speed drops to the "beyond" figure.

FAQ

Can the NVIDIA DGX Spark run AntiFormer?โ–ถ

Yes. At Q4 it generates ~14425.1-17191.5 tok/s and fits up to 1735k context; beyond that it spills to SSD swap and slows to ~528.4-629.7 tok/s.

How much context fits?โ–ถ

At Q4, up to about 1735k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.