removed some gmem pattern mknls with load_ct 0 or 1 since they seem to be...
removed some gmem pattern mknls with load_ct 0 or 1 since they seem to be achieving lower thgoughput than those w/higher load cts
removed some gmem pattern mknls with load_ct 0 or 1 since they seem to be achieving lower thgoughput than those w/higher load cts