summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorLines
2024-02-21STREAM output: Add n_elements_per_dpuBirte Kristina Friesel-2/+2
2023-12-22update gitignoreBirte Kristina Friesel-0/+1
2023-12-15... oopsBirte Kristina Friesel-3/+3
2023-12-15WRAM copy: report latency and throughputBirte Kristina Friesel-20/+33
2023-12-15MRAM-Latency: correctly calculate and label throughputBirte Kristina Friesel-5/+5
2023-12-15nitBirte Kristina Friesel-1/+2
2023-12-15MRAM latency benchmark: report DMA latency and bandwidth; merge r/w modesBirte Kristina Friesel-84/+81
2023-12-12WRAM WiPBirte Kristina Friesel-41/+78
2023-12-11CPU-DPU run scrips: kill stress and mpstatBirte Kristina Friesel-2/+12
2023-12-11Microbenchmarks/CPU-DPU: Extend logfiles with CPU overhead of SDKBirte Kristina Friesel-0/+126
2023-12-08CPU-DPU alloc and transfer microbenchmarks: -search space +cpu loadBirte Kristina Friesel-149/+140
2023-12-05TRNS: updated and extended benchmark scriptsBirte Kristina Friesel-10/+82
2023-12-05TRNS: log number of allocated ranksBirte Kristina Friesel-2/+5
2023-12-01BS: add overhead measurementsBirte Kristina Friesel-46/+144
2023-11-29CPU-DPU: include number of ranks in configuration spaceBirte Kristina Friesel-2/+4
2023-11-28TRNS: Influde free overhead in reconfiguration overheadBirte Kristina Friesel-1/+4
2023-11-27STREAM: Add alloc/load overhead, n_ranks, and n_elementsBirte Kristina Friesel-61/+169
2023-11-27VA: revert prim benchmarks to original configurations; host-specific logsBirte Kristina Friesel-8/+61
2023-11-27VA: add dpuinfo and singlethreaded optionsBirte Kristina Friesel-6/+29
singlethreading does not seem to work at all. Funny SDK.
2023-11-27VA: determine number of ranksBirte Kristina Friesel-0/+3
2023-11-21VA CPU baseline: rename misleading file_size to input_sizeBirte Kristina Friesel-6/+6
2023-11-21VA basilen: … watBirte Kristina Friesel-1/+0
Fix MOpps output
2023-11-21GEMV: Add alloc/load/free measurements and fix a memory leak while at itBirte Kristina Friesel-49/+128
2023-11-21UNI, VA: fix compilation without alloc/load/free benchmarksBirte Kristina Friesel-6/+6
2023-11-20TRNS dpu: Include reconfiguration overhead; fix timer resets in inner loopBirte Kristina Friesel-31/+82
2023-11-20VA: more detail outputBirte Kristina Friesel-7/+33
2023-11-20UNI: Add alloc/load benchmark and upvec exploration scriptBirte Kristina Friesel-44/+147
2023-11-20VA: optionally include dpu_{alloc,load,free} overhead in benchmark outputBirte Kristina Friesel-42/+96
2023-09-25RED: remove stray newlineBirte Kristina Friesel-1/+1
2023-09-21SCAN-SSA: fix dfatool outputBirte Kristina Friesel-13/+9
2023-09-21VA: remove a stray newlineBirte Kristina Friesel-1/+1
2023-09-21CPU-DPU: add microbenchmarks with background loadBirte Kristina Friesel-0/+104
2023-08-03run-opti: PrIM always uses four threads for CPU baselineBirte Kristina Friesel-6/+6
2023-08-01config space and timeout updateBirte Kristina Friesel-7/+15
2023-07-26update config spaceBirte Friesel-33/+44
2023-07-26RED: update benchmark scriptsBirte Friesel-4/+5
2023-07-24HST-S: update config spaceBirte Friesel-6/+7
2023-07-24HST-S CPU Baseline: initialize nr_of_dpus for -x 0Birte Friesel-2/+5
2023-07-24port HST-L to dfatoolBirte Friesel-40/+116
2023-07-24GEMV: update #DPUSBirte Friesel-4/+5
2023-07-24BS: adjust #DPUs for tinosBirte Friesel-2/+3
2023-07-24BFS, BS: remove duplicate configurationsBirte Friesel-3/+3
2023-07-24BFS: adjust strong-full for number of DPUs on tinosBirte Friesel-6/+7
2023-07-24Arithmetic Throughput Microbenchmark: Identify rank/slice/member of faulty DPUsBirte Friesel-3/+23
2023-07-06run-strong-full: add 2544 DPUs (maximum possible on tinos @ UOS)Birte Friesel-12/+30
2023-06-29CPU-DPU: do not scale down data size in BROADCAST benchmarkDaniel Friesel-3/+3
2023-06-29CPU-DPU microbenchmark: more fine-granular state space explorationDaniel Friesel-14/+12
2023-06-22CPU-DPU: measure time for loading binary into dpuDaniel Friesel-30/+152
2023-06-07port cpu-dpu microbenchmark to dfatoolDaniel Friesel-53/+65
2023-06-07RED, SCAN-RSS, SEL, UNI: account for inter-dpu sync in pim kernel timeDaniel Friesel-8/+8