Hello thanks for the quick response,
In the issue I mention that I get Cuda Error in free
when ORT_TENSORRT_MAX_WORKSPACE is too small but it also seems to happen when it is too big.
When I set ORT_TENSORRT_MAX_WORKSPACE=2147483648 on the Jetson TX2 i run into Cuda Error in allocate: 2 (out of memory)
, my tegrastats output is as follows:
RAM 309/7852MB (lfb 1810x4MB) SWAP 0/3926MB (cached 0MB) CPU [1%@345,off,off,0%@345,3%@345,1%@345] EMC_FREQ 2%@1331 GR3D_FREQ 0%@114 APE 150 PLL@40.5C MCPU@40.5C PMIC@100C Tboard@37C GPU@38.5C BCPU@40.5C thermal@39.4C Tdiode@38.25C VDD_SYS_GPU 229/229 VDD_SYS_SOC 765/768 VDD_4V0_WIFI 0/0 VDD_IN 3024/3028 VDD_SYS_CPU 153/156 VDD_SYS_DDR 537/538
RAM 310/7852MB (lfb 1810x4MB) SWAP 0/3926MB (cached 0MB) CPU [1%@345,off,off,0%@345,0%@345,0%@345] EMC_FREQ 2%@1331 GR3D_FREQ 0%@114 APE 150 PLL@40C MCPU@40C PMIC@100C Tboard@37C GPU@38.5C BCPU@40C thermal@39.6C Tdiode@38.25C VDD_SYS_GPU 229/229 VDD_SYS_SOC 765/767 VDD_4V0_WIFI 0/0 VDD_IN 3024/3028 VDD_SYS_CPU 153/155 VDD_SYS_DDR 537/538
RAM 309/7852MB (lfb 1810x4MB) SWAP 0/3926MB (cached 0MB) CPU [1%@345,off,off,0%@345,0%@345,0%@345] EMC_FREQ 2%@1331 GR3D_FREQ 0%@114 APE 150 PLL@40C MCPU@40C PMIC@100C Tboard@37C GPU@38.5C BCPU@40C thermal@39.4C Tdiode@38.25C VDD_SYS_GPU 229/229 VDD_SYS_SOC 765/767 VDD_4V0_WIFI 0/0 VDD_IN 3024/3028 VDD_SYS_CPU 153/155 VDD_SYS_DDR 537/538
RAM 325/7852MB (lfb 1797x4MB) SWAP 0/3926MB (cached 0MB) CPU [40%@652,off,off,3%@652,1%@652,0%@652] EMC_FREQ 2%@1331 GR3D_FREQ 0%@114 APE 150 PLL@40.5C MCPU@40.5C PMIC@100C Tboard@37C GPU@38.5C BCPU@40.5C thermal@39.7C Tdiode@38.25C VDD_SYS_GPU 229/229 VDD_SYS_SOC 841/770 VDD_4V0_WIFI 0/0 VDD_IN 3942/3060 VDD_SYS_CPU 535/169 VDD_SYS_DDR 768/546
RAM 508/7852MB (lfb 1696x4MB) SWAP 0/3926MB (cached 0MB) CPU [23%@1961,off,off,28%@2035,7%@2036,8%@2034] EMC_FREQ 2%@1600 GR3D_FREQ 0%@114 APE 150 PLL@40.5C MCPU@40.5C PMIC@100C Tboard@37C GPU@38.5C BCPU@40.5C thermal@39.9C Tdiode@38.25C VDD_SYS_GPU 229/229 VDD_SYS_SOC 918/775 VDD_4V0_WIFI 0/0 VDD_IN 4324/3104 VDD_SYS_CPU 535/181 VDD_SYS_DDR 902/558
RAM 663/7852MB (lfb 1634x4MB) SWAP 0/3926MB (cached 0MB) CPU [33%@1537,off,off,9%@1727,5%@1728,6%@1727] EMC_FREQ 2%@1600 GR3D_FREQ 0%@114 APE 150 PLL@41C MCPU@41C PMIC@100C Tboard@37C GPU@39C BCPU@41C thermal@40C Tdiode@38.25C VDD_SYS_GPU 229/229 VDD_SYS_SOC 918/780 VDD_4V0_WIFI 0/0 VDD_IN 4248/3142 VDD_SYS_CPU 535/193 VDD_SYS_DDR 902/570
RAM 1032/7852MB (lfb 1527x4MB) SWAP 0/3926MB (cached 0MB) CPU [77%@2036,off,off,2%@2035,0%@2036,0%@2036] EMC_FREQ 5%@1600 GR3D_FREQ 0%@114 APE 150 PLL@41C MCPU@41C PMIC@100C Tboard@37C GPU@39C BCPU@41C thermal@40C Tdiode@38.25C VDD_SYS_GPU 229/229 VDD_SYS_SOC 994/787 VDD_4V0_WIFI 0/0 VDD_IN 4782/3195 VDD_SYS_CPU 917/217 VDD_SYS_DDR 1152/589
RAM 2357/7852MB (lfb 1196x4MB) SWAP 0/3926MB (cached 0MB) CPU [100%@2035,off,off,0%@2035,0%@2037,0%@2036] EMC_FREQ 9%@1600 GR3D_FREQ 0%@114 APE 150 PLL@41C MCPU@41C PMIC@100C Tboard@37C GPU@39C BCPU@41C thermal@40.2C Tdiode@38.5C VDD_SYS_GPU 229/229 VDD_SYS_SOC 994/793 VDD_4V0_WIFI 0/0 VDD_IN 5049/3253 VDD_SYS_CPU 994/241 VDD_SYS_DDR 1324/612
RAM 2358/7852MB (lfb 1185x4MB) SWAP 0/3926MB (cached 0MB) CPU [90%@959,off,off,0%@960,0%@959,0%@960] EMC_FREQ 13%@1331 GR3D_FREQ 0%@114 APE 150 PLL@41C MCPU@41C PMIC@100C Tboard@37C GPU@39C BCPU@41C thermal@40.2C Tdiode@38.5C VDD_SYS_GPU 229/229 VDD_SYS_SOC 994/799 VDD_4V0_WIFI 0/0 VDD_IN 4820/3300 VDD_SYS_CPU 841/259 VDD_SYS_DDR 1171/629
RAM 2389/7852MB (lfb 1148x4MB) SWAP 0/3926MB (cached 0MB) CPU [55%@1728,off,off,1%@1728,0%@1728,0%@1729] EMC_FREQ 6%@1600 GR3D_FREQ 2%@114 APE 150 PLL@41C MCPU@41C PMIC@100C Tboard@37C GPU@39C BCPU@41C thermal@40.2C Tdiode@38.75C VDD_SYS_GPU 229/229 VDD_SYS_SOC 918/803 VDD_4V0_WIFI 0/0 VDD_IN 4324/3330 VDD_SYS_CPU 611/269 VDD_SYS_DDR 902/637
RAM 2448/7852MB (lfb 1103x4MB) SWAP 0/3926MB (cached 0MB) CPU [44%@1970,off,off,14%@2035,0%@2035,1%@2035] EMC_FREQ 5%@1600 GR3D_FREQ 1%@114 APE 150 PLL@41C MCPU@41C PMIC@100C Tboard@37C GPU@39C BCPU@41C thermal@40.2C Tdiode@38.5C VDD_SYS_GPU 229/229 VDD_SYS_SOC 918/806 VDD_4V0_WIFI 0/0 VDD_IN 4363/3360 VDD_SYS_CPU 611/279 VDD_SYS_DDR 921/645
RAM 2508/7852MB (lfb 1059x4MB) SWAP 0/3926MB (cached 0MB) CPU [57%@2034,off,off,3%@2035,0%@2035,0%@2035] EMC_FREQ 4%@1600 GR3D_FREQ 8%@114 APE 150 PLL@41C MCPU@41C PMIC@100C Tboard@37C GPU@39C BCPU@41C thermal@40.2C Tdiode@38.5C VDD_SYS_GPU 229/229 VDD_SYS_SOC 918/809 VDD_4V0_WIFI 0/0 VDD_IN 4324/3387 VDD_SYS_CPU 611/288 VDD_SYS_DDR 921/652
RAM 1492/7852MB (lfb 1237x4MB) SWAP 0/3926MB (cached 0MB) CPU [56%@2033,off,off,21%@2032,5%@2033,0%@2033] EMC_FREQ 4%@1600 GR3D_FREQ 0%@114 APE 150 PLL@41C MCPU@41C PMIC@100C Tboard@37C GPU@39C BCPU@41C thermal@40.2C Tdiode@38.5C VDD_SYS_GPU 229/229 VDD_SYS_SOC 918/812 VDD_4V0_WIFI 0/0 VDD_IN 4286/3411 VDD_SYS_CPU 764/301 VDD_SYS_DDR 902/659
RAM 635/7852MB (lfb 1440x4MB) SWAP 0/3926MB (cached 0MB) CPU [1%@345,off,off,8%@345,12%@345,0%@345] EMC_FREQ 3%@1331 GR3D_FREQ 0%@114 APE 150 PLL@40.5C MCPU@40.5C PMIC@100C Tboard@37C GPU@39C BCPU@40.5C thermal@39.9C Tdiode@38.5C VDD_SYS_GPU 229/229 VDD_SYS_SOC 841/813 VDD_4V0_WIFI 0/0 VDD_IN 3214/3406 VDD_SYS_CPU 306/301 VDD_SYS_DDR 595/657
RAM 636/7852MB (lfb 1440x4MB) SWAP 0/3926MB (cached 0MB) CPU [1%@345,off,off,0%@345,0%@345,0%@345] EMC_FREQ 3%@1331 GR3D_FREQ 0%@114 APE 150 PLL@40.5C MCPU@40.5C PMIC@100C Tboard@37C GPU@39C BCPU@40.5C thermal@39.9C Tdiode@38.25C VDD_SYS_GPU 229/229 VDD_SYS_SOC 765/812 VDD_4V0_WIFI 0/0 VDD_IN 3024/3396 VDD_SYS_CPU 153/297 VDD_SYS_DDR 537/654
RAM 635/7852MB (lfb 1440x4MB) SWAP 0/3926MB (cached 0MB) CPU [0%@345,off,off,0%@345,0%@345,0%@345] EMC_FREQ 3%@1331 GR3D_FREQ 0%@114 APE 150 PLL@40.5C MCPU@40.5C PMIC@100C Tboard@37C GPU@39C BCPU@40.5C thermal@39.9C Tdiode@38.25C VDD_SYS_GPU 229/229 VDD_SYS_SOC 765/810 VDD_4V0_WIFI 0/0 VDD_IN 3024/3387 VDD_SYS_CPU 153/294 VDD_SYS_DDR 537/651
RAM 635/7852MB (lfb 1440x4MB) SWAP 0/3926MB (cached 0MB) CPU [0%@345,off,off,0%@345,0%@345,0%@345] EMC_FREQ 2%@1331 GR3D_FREQ 0%@114 APE 150 PLL@40.5C MCPU@40.5C PMIC@100C Tboard@37C GPU@39C BCPU@40.5C thermal@39.9C Tdiode@38.25C VDD_SYS_GPU 229/229 VDD_SYS_SOC 765/809 VDD_4V0_WIFI 0/0 VDD_IN 3024/3378 VDD_SYS_CPU 153/290 VDD_SYS_DDR 537/649
RAM 635/7852MB (lfb 1440x4MB) SWAP 0/3926MB (cached 0MB) CPU [0%@345,off,off,0%@345,0%@345,0%@345] EMC_FREQ 2%@1331 GR3D_FREQ 0%@114 APE 150 PLL@40.5C MCPU@40.5C PMIC@100C Tboard@37C GPU@38.5C BCPU@40.5C thermal@39.9C Tdiode@38.25C VDD_SYS_GPU 229/229 VDD_SYS_SOC 765/808 VDD_4V0_WIFI 0/0 VDD_IN 3024/3369 VDD_SYS_CPU 153/287 VDD_SYS_DDR 537/646
RAM 635/7852MB (lfb 1440x4MB) SWAP 0/3926MB (cached 0MB) CPU [1%@345,off,off,0%@345,0%@345,0%@345] EMC_FREQ 2%@1331 GR3D_FREQ 0%@114 APE 150 PLL@40.5C MCPU@40.5C PMIC@100C Tboard@37C GPU@38.5C BCPU@40.5C thermal@39.7C Tdiode@38.5C VDD_SYS_GPU 229/229 VDD_SYS_SOC 765/807 VDD_4V0_WIFI 0/0 VDD_IN 3024/3361 VDD_SYS_CPU 153/284 VDD_SYS_DDR 537/643
When I set ORT_TENSORRT_MAX_WORKSPACE=4294967296 on the Jetson TX2 i run into Cuda Error in free: 4 (unspecified launch failure)
, my tegrastats output is as follows:
RAM 311/7852MB (lfb 1809x4MB) SWAP 0/3926MB (cached 0MB) CPU [0%@345,off,off,0%@345,0%@345,0%@345] EMC_FREQ 2%@1331 GR3D_FREQ 0%@114 APE 150 PLL@40.5C MCPU@40.5C PMIC@100C Tboard@37C GPU@38.5C BCPU@40.5C thermal@39.7C Tdiode@38.25C VDD_SYS_GPU 229/229 VDD_SYS_SOC 765/765 VDD_4V0_WIFI 0/0 VDD_IN 3024/3024 VDD_SYS_CPU 153/153 VDD_SYS_DDR 537/537
RAM 328/7852MB (lfb 1792x4MB) SWAP 0/3926MB (cached 0MB) CPU [44%@1428,off,off,1%@1585,2%@1573,4%@1574] EMC_FREQ 1%@1600 GR3D_FREQ 0%@114 APE 150 PLL@40.5C MCPU@40.5C PMIC@100C Tboard@37C GPU@38.5C BCPU@40.5C thermal@39.7C Tdiode@38.25C VDD_SYS_GPU 229/229 VDD_SYS_SOC 841/777 VDD_4V0_WIFI 0/0 VDD_IN 4095/3202 VDD_SYS_CPU 535/216 VDD_SYS_DDR 806/581
RAM 518/7852MB (lfb 1695x4MB) SWAP 0/3926MB (cached 0MB) CPU [41%@960,off,off,9%@499,2%@499,3%@498] EMC_FREQ 3%@1600 GR3D_FREQ 0%@114 APE 150 PLL@40.5C MCPU@40.5C PMIC@100C Tboard@37C GPU@38.5C BCPU@40.5C thermal@39.7C Tdiode@38.5C VDD_SYS_GPU 229/229 VDD_SYS_SOC 918/797 VDD_4V0_WIFI 0/0 VDD_IN 4363/3368 VDD_SYS_CPU 535/262 VDD_SYS_DDR 940/633
RAM 708/7852MB (lfb 1617x4MB) SWAP 0/3926MB (cached 0MB) CPU [42%@2036,off,off,5%@2035,2%@2036,9%@2034] EMC_FREQ 3%@1600 GR3D_FREQ 0%@114 APE 150 PLL@41C MCPU@41C PMIC@100C Tboard@37C GPU@38.5C BCPU@41C thermal@39.7C Tdiode@38.25C VDD_SYS_GPU 229/229 VDD_SYS_SOC 918/812 VDD_4V0_WIFI 0/0 VDD_IN 4286/3483 VDD_SYS_CPU 611/305 VDD_SYS_DDR 902/666
RAM 1233/7852MB (lfb 1476x4MB) SWAP 0/3926MB (cached 0MB) CPU [85%@2034,off,off,1%@2035,0%@2035,0%@2035] EMC_FREQ 6%@1600 GR3D_FREQ 0%@114 APE 150 PLL@41C MCPU@41C PMIC@100C Tboard@37C GPU@38.5C BCPU@41C thermal@40.2C Tdiode@38.5C VDD_SYS_GPU 229/229 VDD_SYS_SOC 994/832 VDD_4V0_WIFI 0/0 VDD_IN 4858/3635 VDD_SYS_CPU 917/373 VDD_SYS_DDR 1190/724
RAM 2394/7852MB (lfb 1186x4MB) SWAP 0/3926MB (cached 0MB) CPU [100%@2036,off,off,0%@2036,0%@2037,0%@2035] EMC_FREQ 10%@1600 GR3D_FREQ 0%@114 APE 150 PLL@41C MCPU@41C PMIC@100C Tboard@37C GPU@39C BCPU@41C thermal@40.3C Tdiode@38.75C VDD_SYS_GPU 229/229 VDD_SYS_SOC 994/849 VDD_4V0_WIFI 0/0 VDD_IN 5126/3784 VDD_SYS_CPU 1070/443 VDD_SYS_DDR 1344/786
RAM 2363/7852MB (lfb 1188x4MB) SWAP 0/3926MB (cached 0MB) CPU [75%@2028,off,off,19%@2031,0%@2032,0%@2033] EMC_FREQ 10%@1600 GR3D_FREQ 0%@114 APE 150 PLL@41C MCPU@41C PMIC@100C Tboard@37C GPU@39C BCPU@41C thermal@40.5C Tdiode@38.75C VDD_SYS_GPU 229/229 VDD_SYS_SOC 917/855 VDD_4V0_WIFI 0/0 VDD_IN 4667/3865 VDD_SYS_CPU 917/486 VDD_SYS_DDR 1094/814
RAM 2400/7852MB (lfb 1141x4MB) SWAP 0/3926MB (cached 0MB) CPU [46%@1966,off,off,4%@2034,1%@2034,0%@2034] EMC_FREQ 7%@1600 GR3D_FREQ 28%@114 APE 150 PLL@41C MCPU@41C PMIC@100C Tboard@37C GPU@39C BCPU@41C thermal@40.2C Tdiode@38.5C VDD_SYS_GPU 229/229 VDD_SYS_SOC 918/860 VDD_4V0_WIFI 0/0 VDD_IN 4324/3903 VDD_SYS_CPU 611/496 VDD_SYS_DDR 921/823
RAM 2452/7852MB (lfb 1098x4MB) SWAP 0/3926MB (cached 0MB) CPU [56%@2034,off,off,1%@2035,1%@2035,0%@2035] EMC_FREQ 5%@1600 GR3D_FREQ 4%@114 APE 150 PLL@41C MCPU@41C PMIC@100C Tboard@37C GPU@39C BCPU@41C thermal@40C Tdiode@38.75C VDD_SYS_GPU 229/229 VDD_SYS_SOC 918/864 VDD_4V0_WIFI 0/0 VDD_IN 4363/3938 VDD_SYS_CPU 611/505 VDD_SYS_DDR 921/831
RAM 2511/7852MB (lfb 1056x4MB) SWAP 0/3926MB (cached 0MB) CPU [34%@1514,off,off,17%@1513,1%@1515,9%@1512] EMC_FREQ 4%@1600 GR3D_FREQ 9%@114 APE 150 PLL@41C MCPU@41C PMIC@100C Tboard@37C GPU@39C BCPU@41C thermal@40.2C Tdiode@38.5C VDD_SYS_GPU 229/229 VDD_SYS_SOC 918/868 VDD_4V0_WIFI 0/0 VDD_IN 4248/3960 VDD_SYS_CPU 535/507 VDD_SYS_DDR 902/836
RAM 2621/7852MB (lfb 1020x4MB) SWAP 0/3926MB (cached 0MB) CPU [28%@345,off,off,0%@345,5%@345,11%@345] EMC_FREQ 5%@1600 GR3D_FREQ 94%@930 APE 150 PLL@41C MCPU@41C PMIC@100C Tboard@37C GPU@41C BCPU@41C thermal@40.2C Tdiode@39.5C VDD_SYS_GPU 1834/336 VDD_SYS_SOC 917/871 VDD_4V0_WIFI 0/0 VDD_IN 5623/4071 VDD_SYS_CPU 382/499 VDD_SYS_DDR 1017/848
RAM 2653/7852MB (lfb 1012x4MB) SWAP 0/3926MB (cached 0MB) CPU [5%@1804,off,off,1%@2034,13%@2029,31%@2034] EMC_FREQ 13%@1600 GR3D_FREQ 99%@1134 APE 150 PLL@41.5C MCPU@41.5C PMIC@100C Tboard@37C GPU@40.5C BCPU@41.5C thermal@41.4C Tdiode@41.75C VDD_SYS_GPU 3437/529 VDD_SYS_SOC 993/879 VDD_4V0_WIFI 0/0 VDD_IN 7418/4280 VDD_SYS_CPU 458/496 VDD_SYS_DDR 1344/879
RAM 710/7852MB (lfb 1439x4MB) SWAP 0/3926MB (cached 0MB) CPU [37%@499,off,off,7%@498,12%@499,22%@498] EMC_FREQ 7%@1600 GR3D_FREQ 0%@1134 APE 150 PLL@41.5C MCPU@41.5C PMIC@100C Tboard@37C GPU@39.5C BCPU@41.5C thermal@41C Tdiode@39.25C VDD_SYS_GPU 382/521 VDD_SYS_SOC 918/881 VDD_4V0_WIFI 0/0 VDD_IN 4248/4278 VDD_SYS_CPU 611/503 VDD_SYS_DDR 921/881
RAM 710/7852MB (lfb 1439x4MB) SWAP 0/3926MB (cached 0MB) CPU [2%@345,off,off,0%@345,0%@345,0%@345] EMC_FREQ 4%@1600 GR3D_FREQ 0%@1134 APE 150 PLL@41C MCPU@41C PMIC@100C Tboard@37C GPU@39C BCPU@41C thermal@40.2C Tdiode@38.75C VDD_SYS_GPU 229/504 VDD_SYS_SOC 841/879 VDD_4V0_WIFI 0/0 VDD_IN 3444/4232 VDD_SYS_CPU 229/488 VDD_SYS_DDR 806/877
RAM 710/7852MB (lfb 1439x4MB) SWAP 0/3926MB (cached 0MB) CPU [0%@345,off,off,0%@345,0%@345,0%@345] EMC_FREQ 3%@1600 GR3D_FREQ 0%@1134 APE 150 PLL@41C MCPU@41C PMIC@100C Tboard@37C GPU@39C BCPU@41C thermal@40.2C Tdiode@38.75C VDD_SYS_GPU 153/486 VDD_SYS_SOC 841/877 VDD_4V0_WIFI 0/0 VDD_IN 3367/4186 VDD_SYS_CPU 153/470 VDD_SYS_DDR 787/872
RAM 710/7852MB (lfb 1439x4MB) SWAP 0/3926MB (cached 0MB) CPU [1%@345,off,off,0%@345,0%@345,0%@345] EMC_FREQ 2%@1600 GR3D_FREQ 0%@1134 APE 150 PLL@40.5C MCPU@40.5C PMIC@100C Tboard@37C GPU@39C BCPU@40.5C thermal@40.2C Tdiode@38.75C VDD_SYS_GPU 153/469 VDD_SYS_SOC 841/875 VDD_4V0_WIFI 0/0 VDD_IN 3367/4145 VDD_SYS_CPU 153/454 VDD_SYS_DDR 787/868
Neither of the tegrastats outputs suggest particularly high memory usage although I guess its quite possible that it is missing the peaks.