Jump to content

DeepComputing

From Wikipedia, the free encyclopedia
The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.
DeepComputing, Ltd.
Company typePrivate
IndustryComputer hardware
RISC-V
Founded2022; 3 years ago (2022)
FounderYuning Liang
Headquarters
ProductsMotherboards, Laptops
Websitedeepcomputing.io

DeepComputing is a technology company based in Hong Kong that develops and sells RISC-V based hardware. Its flagship products are laptops, tablets and PC motherboards with various processors based on 64-bit RISC-V ISA.

History

DeepComputing is founded in 2022, at Sheung Wan, Hong Kong by Yuning Liang. In May, launch of first RISC-V based laptop was rumored by Calista Redmond, former CEO of RISC-V International, with photos of prototypes being shipped to them. In July, DeepComputing together with Xcalibyte (a software company co-founded by Yuning) unveiled DC-ROMA, first commercially available laptop based on a RISC-V based microprocessor.[1][2][3] A month later it was revealed that the SoC inside is an Alibaba TH1520.[4] Later on SoC is changed to be StarFive JH7110. It got released with OpenKylin or Debian Linux pre-installed.

In June 2024, DeepComputing released another laptop named DC-ROMA II, claiming it to be the first RISC-V based laptop to run Ubuntu without any configuration. It was based on SpacemiT K1 SoC, an 8-core CPU running up to 2.0GHz and, together with NPU, 2 TOPS of AI capability.[5][6][7] In August, company announced the first ever RISC-V based tablet named DC ROMA PAD II, which was powered by the same K1 SoC. It was priced around $149.[8]

In November 2024, DeepComputing announced DC-ROMA Mainboard for Framework Laptop 13, a highly modifiable laptop. The partnership was announced back in june by both companies. The motherboard includes a StarFive JH7110 SoC, and was made available for $199.[9][10][11]

In March 2025, DeepComputing announced DC-ROMA AI PC motherboard (initially called Mainboard II)[12], for Framework Laptop 13 again. Its based on ESWIN's EIC7702X SoC that has AI capabilities up to 50 TOPS when NPU (Neural Processing Unit) is enabled. Preorders started from $300, for orders to be effective by july of the same year.[13][14]

Products

Framework Laptop 13 containing DC-ROMA RISC-V AI PC motherboard running Ubuntu 24.04, showing neofetch output

DeepComputing develops hardware ranging from laptops and mainboards to AI PCs and embedded accessories. DeepComputing products only use chips based around RISC-V open standard ISA, which is the main selling point and principle of the company.

DC-ROMA

  • DC-ROMA, first product of the company and world's first commercially available RISC-V laptop. Uses quad-core StarFive JH7110 SoC. It has up to 8GB RAM, 1TB SSD memory, 1080P monocular camera, 2 USB 3.0 ports and Type-C power adaptor. It supports openKylin, Debian and several other Linux distributions.
  • DC-ROMA II, second laptop of the company. Uses octo-core SpacemiT K1 SoC. It has up to 16GB RAM, 1TB SSD memory, 1080P monocular camera, 2 USB 3.0 and Type-C ports, 1 MicroSD port and Type-C power adaptor. It also has GPIO, allowing access to I²C and other types of direct access to the SoC. It comes with Ubuntu or Fedora pre-installed, and supports other Linux distributions.
  • DC-ROMA Pad II, tablet version of DC-ROMA II based on the same K1 SoC.
  • DC-ROMA Mainboard, first Framework compatible motherboard by the company. Uses quad-core JH7110 SoC just like first RISC-V laptop.
  • DC-ROMA AI PC, second Framework motherboard by the company marketed as a local AI-capable PC. Uses octo-core ESWIN EIC7702X SoC. CPU cores are P550 designed by SiFive. SoC contains an NPU that provides AI capability up to 50 TOPS.

Other

  • StationV D300, a workstation based on a RISC-V processor with 64 cores and AI accelerator card support.
  • DC-ROMEO, an RC car based on ESP32-C6

References

  1. ^ "Roma, one of the first RISC-V laptops, may ship in September". Archived from the original on 2025-03-15. Retrieved 2025-07-27.
  2. ^ published, Joel Khalili (2022-07-04). "Take note, Arm: The world's first RISC-V laptop is now available for preorder". TechRadar. Retrieved 2025-07-27.
  3. ^ "RISC-V Laptops Really Arrive in 2025 - IEEE Spectrum". spectrum.ieee.org. Retrieved 2025-07-27.
  4. ^ Dahad, Nitin (2022-10-14). "First RISC-V laptop uses Alibaba TH1520 SoC: embedded.com". Embedded. Retrieved 2025-07-27.
  5. ^ "World's first RISC-V laptop preloaded with Ubuntu". Archived from the original on 2025-06-19. Retrieved 2025-07-28.
  6. ^ published, Jowi Morales (2024-06-13). "The world's first RISC-V laptop gets a big upgrade — DeepComputing doubled the core count, increased clocks to 2 GHz, and added AI capabilities". Tom's Hardware. Retrieved 2025-07-27.
  7. ^ "RISC-V Laptops Really Arrive in 2025 - IEEE Spectrum". spectrum.ieee.org. Retrieved 2025-07-27.
  8. ^ published, Jowi Morales (2024-08-26). "RISC-V CPU arrives on a tablet starting at $149 — DC-ROMA RISC-V Pad II features an octa-core CPU and up to 16GB RAM and 128GB eMMC storage". Tom's Hardware. Retrieved 2025-07-27.
  9. ^ published, Matthew Connatser (2024-11-14). "RISC-V motherboard for Framework 13 laptops and mini-PCs starts at $199 – quad-core RISC-V CPU and 8GB of RAM included". Tom's Hardware. Retrieved 2025-07-27.
  10. ^ "RISC-V main board for Framework laptop shipping". Archived from the original on 2025-04-30. Retrieved 2025-07-27.
  11. ^ "Framework Announces a Partnership with DeepComputing for Its First RISC-V Mainboard". Hackster.io. Retrieved 2025-07-27.
  12. ^ "DeepComputing Unveils the DC-ROMA RISC-V Mainboard II, an AI-Focused Framework 13 Mainboard". Hackster.io. Retrieved 2025-07-28.
  13. ^ published, Sunny Grimm (2025-03-10). "RISC-V mini AI PC that fits inside a Framework laptop shell revealed — DeepComputing's DC-ROMA RISC-V AI PC claims 50 TOPS, 64GB RAM". Tom's Hardware. Retrieved 2025-07-27.
  14. ^ "DeepComputing Unwraps RISC-V AI PC Delivering 50 TOPS Performance - News". www.allaboutcircuits.com. Retrieved 2025-07-28.