deeprec win AMD Ryzen 9 7950X 16-Core testing with a ASUS ROG CROSSHAIR X670E HERO (1101 BIOS) and AMD Radeon RX 7900 XTX 24GB on Ubuntu 22.04 via the Phoronix Test Suite.
HTML result view exported from: https://openbenchmarking.org/result/2304235-PTS-DEEPRECW27&grt .
deeprec win Processor Motherboard Chipset Memory Disk Graphics Audio Monitor Network OS Kernel Desktop Display Server OpenGL Vulkan Compiler File-System Screen Resolution a b c AMD Ryzen 9 7950X 16-Core @ 4.50GHz (16 Cores / 32 Threads) ASUS ROG CROSSHAIR X670E HERO (1101 BIOS) AMD Device 14d8 32GB 2048GB SOLIDIGM SSDPFKKW020X7 + 2000GB AMD Radeon RX 7900 XTX 24GB (2304/1249MHz) AMD Device ab30 ASUS MG28U Intel I225-V + Intel Wi-Fi 6 AX210/AX211/AX411 Ubuntu 22.04 6.3.0-060300rc7daily20230417-generic (x86_64) GNOME Shell 42.5 X Server 1.21.1.3 + Wayland 4.6 Mesa 23.2.0-devel (git-f6fb189 2023-04-18 jammy-oibaf-ppa) (LLVM 15.0.7 DRM 3.52) 1.3.246 GCC 11.3.0 ext4 3840x2160 OpenBenchmarking.org Kernel Details - Transparent Huge Pages: madvise Processor Details - Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0xa601203 Python Details - Python 3.10.9 Security Details - itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
deeprec win deeprec: BST - BF16 deeprec: BST - BF16 deeprec: BST - FP32 deeprec: BST - FP32 deeprec: DIN - BF16 deeprec: DIN - BF16 deeprec: DIN - FP32 deeprec: DIN - FP32 deeprec: PLE - BF16 deeprec: PLE - BF16 deeprec: PLE - FP32 deeprec: PLE - FP32 deeprec: DLRM - BF16 deeprec: DLRM - BF16 deeprec: DLRM - FP32 deeprec: DLRM - FP32 deeprec: MMOE - BF16 deeprec: MMOE - BF16 deeprec: MMOE - FP32 deeprec: MMOE - FP32 deeprec: DCNv2 - BF16 deeprec: DCNv2 - BF16 deeprec: DCNv2 - FP32 deeprec: DCNv2 - FP32 a b c 5442.76 10.63 10459.37 20.43 39620.03 77.39 45933.94 89.71 19539.80 38.16 34192.93 66.78 40034.28 78.19 153853.71 300.50 24768.41 48.38 98473.37 192.33 10981.04 21.45 15386.73 30.05 5390.11 10.53 10469.73 20.45 39371.49 76.90 46189.13 90.21 19521.64 38.13 33588.20 65.60 39866.20 77.86 154485.76 301.73 24744.99 48.33 99439.16 194.22 11010.90 21.51 15394.56 30.07 5397.29 10.54 10450.02 20.41 39664.62 77.47 46184.24 90.20 19562.81 38.21 33755.33 65.93 39957.45 78.04 153879.88 300.55 24737.78 48.32 99015.53 193.39 10997.10 21.48 15407.15 30.09 OpenBenchmarking.org
DeepRec Model: BST - Data Type: BF16 OpenBenchmarking.org Throughput, More Is Better DeepRec Model: BST - Data Type: BF16 a b c 1200 2400 3600 4800 6000 SE +/- 75.87, N = 3 SE +/- 5.26, N = 3 SE +/- 4.97, N = 3 5442.76 5390.11 5397.29
DeepRec Model: BST - Data Type: BF16 OpenBenchmarking.org Gstep / sec, More Is Better DeepRec Model: BST - Data Type: BF16 a b c 3 6 9 12 15 SE +/- 0.15, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 10.63 10.53 10.54
DeepRec Model: BST - Data Type: FP32 OpenBenchmarking.org Throughput, More Is Better DeepRec Model: BST - Data Type: FP32 a b c 2K 4K 6K 8K 10K SE +/- 10.62, N = 3 SE +/- 15.53, N = 3 SE +/- 7.31, N = 3 10459.37 10469.73 10450.02
DeepRec Model: BST - Data Type: FP32 OpenBenchmarking.org Gstep / sec, More Is Better DeepRec Model: BST - Data Type: FP32 a b c 5 10 15 20 25 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 20.43 20.45 20.41
DeepRec Model: DIN - Data Type: BF16 OpenBenchmarking.org Throughput, More Is Better DeepRec Model: DIN - Data Type: BF16 a b c 8K 16K 24K 32K 40K SE +/- 59.19, N = 3 SE +/- 283.43, N = 3 SE +/- 94.02, N = 3 39620.03 39371.49 39664.62
DeepRec Model: DIN - Data Type: BF16 OpenBenchmarking.org Gstep / sec, More Is Better DeepRec Model: DIN - Data Type: BF16 a b c 20 40 60 80 100 SE +/- 0.12, N = 3 SE +/- 0.55, N = 3 SE +/- 0.18, N = 3 77.39 76.90 77.47
DeepRec Model: DIN - Data Type: FP32 OpenBenchmarking.org Throughput, More Is Better DeepRec Model: DIN - Data Type: FP32 a b c 10K 20K 30K 40K 50K SE +/- 235.44, N = 3 SE +/- 226.20, N = 3 SE +/- 158.55, N = 3 45933.94 46189.13 46184.24
DeepRec Model: DIN - Data Type: FP32 OpenBenchmarking.org Gstep / sec, More Is Better DeepRec Model: DIN - Data Type: FP32 a b c 20 40 60 80 100 SE +/- 0.46, N = 3 SE +/- 0.44, N = 3 SE +/- 0.31, N = 3 89.71 90.21 90.20
DeepRec Model: PLE - Data Type: BF16 OpenBenchmarking.org Throughput, More Is Better DeepRec Model: PLE - Data Type: BF16 a b c 4K 8K 12K 16K 20K SE +/- 5.68, N = 3 SE +/- 86.66, N = 3 SE +/- 75.93, N = 3 19539.80 19521.64 19562.81
DeepRec Model: PLE - Data Type: BF16 OpenBenchmarking.org Gstep / sec, More Is Better DeepRec Model: PLE - Data Type: BF16 a b c 9 18 27 36 45 SE +/- 0.01, N = 3 SE +/- 0.17, N = 3 SE +/- 0.15, N = 3 38.16 38.13 38.21
DeepRec Model: PLE - Data Type: FP32 OpenBenchmarking.org Throughput, More Is Better DeepRec Model: PLE - Data Type: FP32 a b c 7K 14K 21K 28K 35K SE +/- 81.30, N = 3 SE +/- 302.32, N = 9 SE +/- 358.30, N = 5 34192.93 33588.20 33755.33
DeepRec Model: PLE - Data Type: FP32 OpenBenchmarking.org Gstep / sec, More Is Better DeepRec Model: PLE - Data Type: FP32 a b c 15 30 45 60 75 SE +/- 0.16, N = 3 SE +/- 0.59, N = 9 SE +/- 0.70, N = 5 66.78 65.60 65.93
DeepRec Model: DLRM - Data Type: BF16 OpenBenchmarking.org Throughput, More Is Better DeepRec Model: DLRM - Data Type: BF16 a b c 9K 18K 27K 36K 45K SE +/- 101.42, N = 3 SE +/- 46.34, N = 3 SE +/- 88.98, N = 3 40034.28 39866.20 39957.45
DeepRec Model: DLRM - Data Type: BF16 OpenBenchmarking.org Gstep / sec, More Is Better DeepRec Model: DLRM - Data Type: BF16 a b c 20 40 60 80 100 SE +/- 0.20, N = 3 SE +/- 0.09, N = 3 SE +/- 0.17, N = 3 78.19 77.86 78.04
DeepRec Model: DLRM - Data Type: FP32 OpenBenchmarking.org Throughput, More Is Better DeepRec Model: DLRM - Data Type: FP32 a b c 30K 60K 90K 120K 150K SE +/- 554.46, N = 3 SE +/- 611.96, N = 3 SE +/- 661.29, N = 3 153853.71 154485.76 153879.88
DeepRec Model: DLRM - Data Type: FP32 OpenBenchmarking.org Gstep / sec, More Is Better DeepRec Model: DLRM - Data Type: FP32 a b c 70 140 210 280 350 SE +/- 1.08, N = 3 SE +/- 1.20, N = 3 SE +/- 1.29, N = 3 300.50 301.73 300.55
DeepRec Model: MMOE - Data Type: BF16 OpenBenchmarking.org Throughput, More Is Better DeepRec Model: MMOE - Data Type: BF16 a b c 5K 10K 15K 20K 25K SE +/- 8.20, N = 3 SE +/- 17.25, N = 3 SE +/- 16.28, N = 3 24768.41 24744.99 24737.78
DeepRec Model: MMOE - Data Type: BF16 OpenBenchmarking.org Gstep / sec, More Is Better DeepRec Model: MMOE - Data Type: BF16 a b c 11 22 33 44 55 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 48.38 48.33 48.32
DeepRec Model: MMOE - Data Type: FP32 OpenBenchmarking.org Throughput, More Is Better DeepRec Model: MMOE - Data Type: FP32 a b c 20K 40K 60K 80K 100K SE +/- 607.13, N = 3 SE +/- 189.62, N = 3 SE +/- 204.46, N = 3 98473.37 99439.16 99015.53
DeepRec Model: MMOE - Data Type: FP32 OpenBenchmarking.org Gstep / sec, More Is Better DeepRec Model: MMOE - Data Type: FP32 a b c 40 80 120 160 200 SE +/- 1.19, N = 3 SE +/- 0.37, N = 3 SE +/- 0.40, N = 3 192.33 194.22 193.39
DeepRec Model: DCNv2 - Data Type: BF16 OpenBenchmarking.org Throughput, More Is Better DeepRec Model: DCNv2 - Data Type: BF16 a b c 2K 4K 6K 8K 10K SE +/- 29.29, N = 3 SE +/- 11.07, N = 3 SE +/- 14.84, N = 3 10981.04 11010.90 10997.10
DeepRec Model: DCNv2 - Data Type: BF16 OpenBenchmarking.org Gstep / sec, More Is Better DeepRec Model: DCNv2 - Data Type: BF16 a b c 5 10 15 20 25 SE +/- 0.06, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 21.45 21.51 21.48
DeepRec Model: DCNv2 - Data Type: FP32 OpenBenchmarking.org Throughput, More Is Better DeepRec Model: DCNv2 - Data Type: FP32 a b c 3K 6K 9K 12K 15K SE +/- 19.76, N = 3 SE +/- 11.07, N = 3 SE +/- 13.19, N = 3 15386.73 15394.56 15407.15
DeepRec Model: DCNv2 - Data Type: FP32 OpenBenchmarking.org Gstep / sec, More Is Better DeepRec Model: DCNv2 - Data Type: FP32 a b c 7 14 21 28 35 SE +/- 0.04, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 30.05 30.07 30.09
Phoronix Test Suite v10.8.5