Geerling llama.cpp Vulkan on rtx2080ti (pi host): llama-2-13b Q4_K_M pp512 1350.00 tok/s
1350.0 tok/s (pp512)
Fields
id
benchmarks/geerling-vulkan-rtx2080ti-pi-llama-2-13b-q4km-pp512Metric
prefill_tok_s
Value
1,350
Unit
tok/s (pp512)
Model
llama-2-13b.Q4_K_M.gguf (7.33 GiB, 13.02 B params)
Methodology
THIRD-PARTY MEASURED (Jeff Geerling, geerlingguy/ai-benchmarks). llama.cpp build f9ec8858e (7474), backend Vulkan, -fa 1 -mmp 0 -ngl 99 (flash attention on, mmap off). Host: Raspberry Pi CM5 16GB, Pi OS Trixie, PCIe-attached card. Model file size 7.33 GiB, 13.02 B params, Q4_K_M. prefill tok/s, prompt length 512, per-request, batch 1. NOT aggregate. Whole-system wall power during the run 83.1 W (node scope, not device).
Measured by
third_party
Reproducible
yes
Mean power (W)
83.1
Power scope
node
Thermal state
unknown
notes
Source issue prints the llama-bench mean and the +/- spread; the stored value is the mean. The Pi-host rows are the same physical GPU on a PCIe-constrained host, so they are not a card ceiling.
Notes
Source issue prints the llama-bench mean and the +/- spread; the stored value is the mean. The Pi-host rows are the same physical GPU on a PCIe-constrained host, so they are not a card ceiling.