Comment Comment by girvo > DGX Spark-alike is really just asking for trouble. Prefill kills perf.You're right that prefill kills perf, but shrug the GB10 has far more compute than it has memory bandwidth, so prefill isn't it's bottleneck. parent girvo 4 months ago
You're right that prefill kills perf, but shrug the GB10 has far more compute than it has memory bandwidth, so prefill isn't it's bottleneck.