AMD Versal AI Edge VEK280 (VE2802) vs Rockchip RK3588 (NPU, single core of 3)

FPGA / Versal coverage note: 1,243 rows come from the AMD Vitis AI 3.0/3.5 Model Zoo tables; 733 (59%) are image classification. Detection, segmentation and other workloads are thinner, and there is no LLM or speech data for these boards. All are vendor-published (grade D). FPGA / Versal overview.

AMDeval kit (adaptive SoC)No licensed photo yet

Close-up of a Rockchip RK3588 system-on-chip package on a circuit board — Rockchip RK3588 SoC package. Photo: Popolon (Commons uploader; CC0, attribution not required), CC0, via Wikimedia Commons.

AMD Versal AI Edge VEK280 (VE2802) is faster in the only like-for-like comparison (largest gap: image classification, 3.00×).

	AMD Versal AI Edge VEK280 (VE2802)	Rockchip RK3588 (NPU, single core of 3)
Best image classification	8598 FPS D Older data (2023)	459 FPS C 2026-10-13 (page last edited)
Power	n/a	n/a
Perf / W	n/a	n/a
Price	n/a	n/a
Workloads with data	4 of 7	5 of 7

Best published image classification result per platform (may be from different like-for-like groups — see data). n/a = not published; never estimated.

Show data — all workloads, sources and like-for-like groups

object detection

⚠ MIXED COMPARISON — not like-for-like. Baseline small (YOLO-s) at batch 1. Differs: model-size class mobile-class CNN vs small (YOLO-s); batch unknown vs 1.

Platform	Result	Model · precision · stack	Grade	Power	Source
Rockchip RK3588 (NPU, single core of 3)	66.50 FPS throughput · b=1	yolov5s_relu INT8 · RKNN-Toolkit2 / RKNPU (rknn_model_zoo) · small (YOLO-s)	C	unknown power n/a no power published	Rockchip rknn_model_zoo README benchmark table 2026-10-13 (page last edited)
AMD Versal AI Edge VEK280 (VE2802) MIXED: model-size class mobile-class CNN vs small (YOLO-s); batch unknown vs 1	1933 FPS E2E throughput, multi thread · b=unknown	tf_ssdmobilenetv1_3.5 INT8 · Vitis AI 3.5 (VART) · mobile-class CNN	D	unknown power n/a no power published	Vitis AI 3.5 Model Zoo performance table Older data (2023)

No shared like-for-like group exists for this pair.

instance segmentation

Only Rockchip RK3588 (NPU, single core of 3) has published instance segmentation results.

Platform	Result	Model · precision · stack	Grade	Power	Source
Rockchip RK3588 (NPU, single core of 3)	71.50 FPS throughput · b=1	yolov5n-seg INT8 · RKNN-Toolkit2 / RKNPU (rknn_model_zoo) · nano (YOLO-n)	C	unknown power n/a no power published	Rockchip rknn_model_zoo README benchmark table 2026-10-13 (page last edited)

AMD Versal AI Edge VEK280 (VE2802):

◌

No published instance segmentation results for AMD Versal AI Edge VEK280 (VE2802) or Rockchip RK3588 (NPU, single core of 3) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.

pose estimation

⚠ MIXED COMPARISON — not like-for-like. Baseline nano (YOLO-n) at batch 1. Differs: model-size class mobile-class CNN vs nano (YOLO-n); batch unknown vs 1.

Platform	Result	Model · precision · stack	Grade	Power	Source
Rockchip RK3588 (NPU, single core of 3)	58.30 FPS throughput · b=1	yolov8n-pose INT8 · RKNN-Toolkit2 / RKNPU (rknn_model_zoo) · nano (YOLO-n)	C	unknown power n/a no power published	Rockchip rknn_model_zoo README benchmark table 2026-10-13 (page last edited)
AMD Versal AI Edge VEK280 (VE2802) MIXED: model-size class mobile-class CNN vs nano (YOLO-n); batch unknown vs 1	428 FPS E2E throughput, multi thread · b=unknown	pt_movenet_3.5 INT8 · Vitis AI 3.5 (VART) · mobile-class CNN	D	unknown power n/a no power published	Vitis AI 3.5 Model Zoo performance table Older data (2023)

No shared like-for-like group exists for this pair.

image classification

✓ Like-for-like: model-size class mid-size CNN, batch 1. Model, precision and stack may still differ.

Platform	Result	Model · precision · stack	Grade	Power	Source
AMD Versal AI Edge VEK280 (VE2802)	337 FPS throughput · b=1	resnet50 (TRD example xmodel) INT8 · Vitis AI 3.5 DPUCV2DX8G reference design, test_dpu_runner_mt · mid-size CNN	D	unknown power n/a no power published	Vitis AI 3.5 VEK280 DPUCV2DX8G reference design (ResNet50 example) Older data (2023)
Rockchip RK3588 (NPU, single core of 3)	113 FPS throughput · b=1	resnet50-v2-7 INT8 · RKNN-Toolkit2 / RKNPU (rknn_model_zoo) · mid-size CNN	C	unknown power n/a no power published	Rockchip rknn_model_zoo README benchmark table 2026-10-13 (page last edited)

semantic segmentation

⚠ MIXED COMPARISON — not like-for-like. Baseline mobile-class CNN at batch 1. Differs: model-size class mid-size CNN vs mobile-class CNN; batch unknown vs 1.

Platform	Result	Model · precision · stack	Grade	Power	Source
Rockchip RK3588 (NPU, single core of 3)	34.70 FPS throughput · b=1	deeplab-v3-plus-mobilenet-v2 INT8 · RKNN-Toolkit2 / RKNPU (rknn_model_zoo) · mobile-class CNN	C	unknown power n/a no power published	Rockchip rknn_model_zoo README benchmark table 2026-10-13 (page last edited)
AMD Versal AI Edge VEK280 (VE2802) MIXED: model-size class mid-size CNN vs mobile-class CNN; batch unknown vs 1	1564 FPS E2E throughput, multi thread · b=unknown	tf2_2D-UNET_3.5 INT8 · Vitis AI 3.5 (VART) · mid-size CNN	D	unknown power n/a no power published	Vitis AI 3.5 Model Zoo performance table Older data (2023)

No shared like-for-like group exists for this pair.

speech-to-text

◌

No published speech-to-text results for AMD Versal AI Edge VEK280 (VE2802) or Rockchip RK3588 (NPU, single core of 3) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.

LLM generation

◌

No published LLM generation results for AMD Versal AI Edge VEK280 (VE2802) or Rockchip RK3588 (NPU, single core of 3) yet.
We only show cited public figures — nothing is estimated to fill gaps. Requests tell us what to source or test next.

How to read these numbers

Every number is labelled. Confidence: A MLPerf closed, audited · B documented third-party method · C community / review · D vendor-published, not independently verified · E estimate. Power basis: measured with the result · spec vendor spec/envelope · power n/a not published.

Machine-readable: /find.json (see llms.txt); full tables are in “Show data”. Confidence grades: A audited MLPerf closed · B documented third-party · C community · D vendor-published · E estimate. Unknown values are shown as unknown, never estimated. Generated 2026-10-16 from public sources.