Turing Machines

Turing RK1 Compute Module

Expert rating7.5 / 10

SoC:: Rockchip RK3588
CPU:: Octa-core (4x Cortex-A76 + 4x Cortex-A55)
RAM:: Up to 32GB LPDDR4x
Form Factor:: Turing Pi 2 compute module
Price:: From $110

Consumercarts is reader-supported. Learn more

Disclaimer: The price and specification of the given product may differ from the actual product. Please confirm on the retailer's website before buying.

Manufacturer: Turing Machines Last updated: September 11, 2026

Turing RK1 Compute Module Review

The Turing RK1 is a compute module built specifically to slot into the Turing Pi 2 cluster board already on this site, giving buyers of that board an RK3588-class compute option alongside Raspberry Pi CM4 and NVIDIA Jetson modules.

Rather than a general-purpose board, the Turing RK1 is designed from the ground up to populate one of the Turing Pi 2's four compute module slots, letting buyers mix and match RK1, CM4, and Jetson modules in a single cluster.

Purpose-Built for the Turing Pi 2

RK3588 Performance in Cluster Form

The same class of Rockchip RK3588 silicon used in several standalone boards already on this site (including the Banana Pi BPI-M7 and Khadas Edge2) here takes the form of a swappable cluster node, aimed at homelab Kubernetes and distributed-computing experimentation.

Comparison to Alternatives

Against the Turing Pi 2 already on this site, the RK1 is the compute module that populates it rather than the cluster board itself — the two are typically purchased together. Against the Radxa CM5 already on this site, the Turing RK1 targets cluster-board deployment specifically, while the CM5 targets Pi CM4-compatible embedded product carrier boards.

Our verdict

Reviewed by Burooj Alam. Rated by us, not by users.

Expert rating7.5 / 10

RK3588-class compute in a swappable cluster node, useful only with a Turing Pi 2 carrier.

Best for

Turing Pi 2 owners who want RK3588-class compute nodes for homelab cluster experiments.

Not for

Anyone without a Turing Pi 2 (or compatible) carrier board to plug it into.

Specifications

What these specs mean

Identity

Brand	Turing Machines

Price	110 USD

Measured results

Board measurements

Memory copy bandwidth (tinymembench memcpy)	12,579.7 MB/s	Board as reviewed · Turing RK1 (review 38) · Source
Idle power	2.85 W	Board as reviewed · Turing RK1 (review 38) · Source
Power under full CPU load (stress-ng)	15 W	Board as reviewed · Turing RK1 (review 38) · Source

Measured by the author on one board. Power is measured at the wall for the whole board. Memory bandwidth is the tinymembench memcpy result.

Retrieved 23 Sept 2026

Jeff Geerling, sbc-reviews (github.com/geerlingguy/sbc-reviews, web/data/results.yml). Copyright (c) 2023 Jeff Geerling. Used under the MIT License; full permission notice: github.com/geerlingguy/sbc-reviews/blob/master/LICENSE

FAQs

What are the key specifications of the Turing RK1 Compute Module (Compute Module/Rockchip RK3588/Octa-Core/Up to 32GB/Turing Pi 2 Cluster Node)?

The Turing RK1 Compute Module (Compute Module/Rockchip RK3588/Octa-Core/Up to 32GB/Turing Pi 2 Cluster Node) is a single board computer from Turing Machines. Key specs: SoC Rockchip RK3588, CPU Octa-core (4x Cortex-A76 + 4x Cortex-A55), RAM Up to 32GB LPDDR4x, Price From $110.

Will the Turing RK1 Compute Module (Compute Module/Rockchip RK3588/Octa-Core/Up to 32GB/Turing Pi 2 Cluster Node) work with my current setup?

The Turing RK1 Compute Module (Compute Module/Rockchip RK3588/Octa-Core/Up to 32GB/Turing Pi 2 Cluster Node) lists SoC is Rockchip RK3588. Check that against the hardware you already own before buying, since this is the field that most often decides whether a part fits.

Who is the Turing RK1 Compute Module (Compute Module/Rockchip RK3588/Octa-Core/Up to 32GB/Turing Pi 2 Cluster Node) best suited to?

Best suited to: Turing Pi 2 owners who want RK3588-class compute nodes for homelab cluster experiments.

What are the drawbacks of the Turing RK1 Compute Module (Compute Module/Rockchip RK3588/Octa-Core/Up to 32GB/Turing Pi 2 Cluster Node)?

Not the best fit for: Anyone without a Turing Pi 2 (or compatible) carrier board to plug it into.