Posts with this tag: 8
GPU idle was 60 percent of decode wall time, and the largest share was host-to-device copies running at 3.1 to 3.4 GB/s against 28.5 GB/s on a healthy boot. The root port negotiated 2.5 GT/s. Several timed A/Bs from that boot are void, one fusion ratio survives, and the machine is still on the slow link.
ReadStd.math exp disagreed between the CPU and the GPU on 19,437 of 40,017 inputs. A shared exp written with explicit fma matched on all 40,017, at 3.6 times the cost in an exp-bound kernel. One attention layer then matched bit for bit, and a float Freivalds certificate caught every planted fault.
ReadThe fused q4x kernels cut 1244 dispatches per decode token to 992, behind a gate that had never been red. One nudge of about 5e-7 per kernel turned it red three times, each time naming its own kernel. The identity check beside it has still never been seen to fail.
ReadFour days of a wrong verdict ("unusable with our Mojo kernels") ended when we unbound the integrated GPU. A plain kernel then read SQ_WAVES 32,768 exactly, and a second fault, counters reading zero, came out from behind the first.
ReadMojo has no ray-tracing API, yet a 20-line probe emits image_bvh_intersect_ray for gfx1100. The probe is compile only, with an all-zero descriptor and one component set, so it proves reachability and nothing more.
Read40.0 GB read and 15.9 GB written per launch is 59 to 87 percent of the card's bandwidth, depending on which clock mode you trust. By the review's estimate the biggest items are repeated reads of factors and vectors, not the Jacobian we blamed.
ReadA day with 135 GPU jobs and 34 failures, none of which needed the GPU to find. The fix is not a faster GPU, it's treating the card as a shared, budgeted resource with a preflight step before anything touches it.
ReadSquare GEMM is the benchmark everyone runs; decode is M=1 streaming weights out of HBM. Four frozen rounds, three wrong predictions, and a one-percent win that came from occupancy relief rather than fewer bytes.
Read
