Google GF180MCU open shuttle (GFMPW‑0)Run by Google and Efabless

40 public designs on one GF180MCU reticle

What was built

The run as a whole: what its designs add up to.

One of the free open-source GF180MCU shuttles that Google sponsored and Efabless ran. Each design sat in the user area of the Caravel harness, which took 55% of every die for its management SoC, pads and wiring. Its designs are measured here the same way as wafer.space's own runs, so you can compare them with a wafer.space die, where everything inside the pad ring is your design. The other Google shuttle measured here is MPW-18H1.

40

Public designs

filling all 40 slots

40

Open source

with a public repository

8.4M

Transistors

across the unique designs

192k

Logic cells

gates and flip-flops

256 bytes

SRAM

in foundry and custom macros

40

Whole dies

every slot a whole die

Private designs on the reticle are not in the public layout and are not counted. Designs placed more than once are counted once.

Make a chip of your own

These designs were made free on one of Google's shuttles. The same GF180MCU process is open to anyone on wafer.space: start small with a Tiny Tapeout tile, or go straight to a whole die on the next wafer.space run.

Make a whole chip of your own

Book a die to itself on wafer.space Run 3: orders close 9 December 2026.

Reserve a slot
Or start with a Tiny Tapeout tile

Tiny Tapeout GF 26c and 26d take projects until 7 December 2026, to be made on wafer.space Run 3.

Submit a project
Get a Tiny Tapeout chip

Every Tiny Tapeout chip comes on a board you can plug in, with all of its projects ready to try.

Buy a chip and board

The reticle

Every design rendered from the layout at the same scale and placed where it sits on the 32 × 26 mm reticle. The slots are named as in the shuttle's own list: rows lettered from the top, columns numbered from the right. Point at a die to see its name, and select it for its details.Every die is listed further down.

Slots

How much of the reticle each design took: a 1 × 1 slot is a whole die of its own.

1 × 140
Total40

Layout on GitHub

Shuttle status

How this run went, from the call for designs to the finished chips.

  • 31 October 2022 Submissions open Google and Efabless open the first GF180MCU Open MPW shuttle, free to open source designs
  • 5 December 2022, 11:59 PT Tapeout deadline Final deadline to submit projects; 86 projects submitted for 40 slots
  • 21 December 2022 Final layouts Efabless commits the final layouts of the 40 selected projects, each inside the Caravel harness chip
  • 3 January 2023 Layouts updated Last layout updates by Efabless (PR boundary layer removed)
  • 31 March 2023 Assembly Packaging and assembly, on Efabless's schedule
  • 2 June 2023 Delivery Chips and evaluation boards delivered to designers, on Efabless's schedule
  • Today, 24 October 2026

From the Efabless GFMPW-0 shuttle page (archived) and the shuttle repositories.

How people used their silicon

Where the area of 40 dies (805 mm² of silicon) actually went

40%
55%
  • Logic cells 1% — gates and flip-flops doing the design's actual work
  • Filler and tap cells 4% — inserted by the tools to keep rows continuous and wells tied
  • Other 40% — routing channels, analog and custom layout, power grid and empty space
  • Outside the user area 55% — the Caravel harness around the user's design: its management SoC, pads and wiring

Standard cell libraries

The ready-made gate libraries the designs were built from: how many tracks tall each cell is, and the voltage it runs at.

Standard cell libraryLogic cellsShare
7t-5V gf180mcu_fd_sc_mcu7t5v0 192k 100%

Area budget of every design

Each bar is one die, split the same way as above. Sorted by how much of the die is logic.

Show all 40 dies

Which parts of the process were used

Transistor voltages and optional devices across the 40 designs

Transistor voltages

5 V 38
6 V 2

A design counts for a voltage when its standard cells, SRAM or its own transistors use it. 5 V and 6 V transistors share the thick gate oxide; a marker layer in the layout tells them apart.

Optional devices drawn by designers

DeviceDesignsWhich
eFuses1 Ophelia eFPGA

Not drawn by any design: Native transistors, 10 V LDMOS, MIM capacitors, MOS capacitors, High-resistance poly resistors, Unsalicided poly resistors, Salicided poly resistors, Diffusion resistors, N-well resistors, Metal resistors, Diodes, Well diodes, Deep N-well, Bipolar transistors, Schottky diodes, Zener diodes, One-time programmable memory, Inductors.

Only devices outside standard cells, SRAM macros and I/O pads count, since every die has those. A single test structure counts the same as a large block.

What fits on 1 mm² of GF180MCU silicon?

Theoretical maximum against what designers achieved

The maximum packs one cell type edge to edge with no room for wiring. Real designs need the wiring. Each box gives the most of a cell that fits with any GF180MCU library, then with the libraries this run used.

~114k

Inverters per mm²

GF180MCU maximum,
with AS 7t-3.3V inv_2

With this run's libraries:

7t-5V: 114k inv_1

~91k

NAND gates per mm²

GF180MCU maximum,
with 7t-5V nand2_1

With this run's libraries:

7t-5V: 91k nand2_1

~76k

Buffers per mm²

GF180MCU maximum,
with AS 7t-3.3V buff_2

With this run's libraries:

7t-5V: 76k buf_1

~17k

Flip-flops per mm²

GF180MCU maximum,
with AS 7t-3.3V dfxtp_2

With this run's libraries:

7t-5V: 16k dffq_1

7k

Best core average

logic cells/mm² of core — 9% of the buffer maximum

Ophelia eFPGA

19k

Densest mm²

logic cells — 25% of the buffer maximum

Ophelia eFPGA

<1k

Typical die

median logic cells/mm² of core across 16 digital dies (1,000 or more logic cells)

1% of the core covered by logic cells

Percentages compare with the buffer maximum for each design's own library: buffers are among the smallest and most common logic cells, so a design made only of buffers, packed edge to edge with no wiring, would reach 100%.

Sizing your own design

What each slot size held on this run, and what it would hold at its densities

Logic cells in its digital designs1,000 or more logic cells Logic cells at this run's density Theoretical maximum, 7t-5V cells
Slot Die  Core area Designs  Most Median  Typical<1k / mm²Best7k / mm² Densest19k / mm² NAND gatesnand2_1Flip-flopsdffq_1
1 × 13.93 × 5.12 mm 9.0 mm²40 65k4k 4k 65k 169k 820k 141k

Swipe the table sideways for more columns.

Core area is the area of each die's user_project_wrapper cell. "Typical" and "best" multiply the core area by this run's median and best core densities, and "densest" by the densest 1 mm² any design on this run reached; SRAM and analog blocks take their area out of that. The theoretical maxima fill the core with nothing but that one cell of 7t-5V, the library most used on this run, edge to edge with no routing.

Pick the slot size your design needs on wafer.space Run 3: orders close 9 December 2026.

Logic density of every digital design

A digital die here is one with 1,000 or more logic cells. Solid bar: average over the core. Outlined bar: the densest mm² outside SRAM (a dash where memory covers most of the design). The scale ends at the buffer theoretical maximum (76k cells/mm²). The numbers are core / densest, in logic cells per mm².

Show all 16 digital dies
Every digital die's densities as a table
DesignSlotLibraryLogic cells Inside core
per mm²
Densest mm²
per mm²
Core coveredCore
mm²
Ophelia eFPGA 45D6 1 × 17t-5V 64,755 7,197 18,753 17% 9.00
One-hot FABulous FPGA 56C6 1 × 17t-5V 32,037 3,561 5,395 11% 9.00
VerilogBoy C979 1 × 17t-5V 27,097 3,012 18,124 8% 9.00
Serv tri-core (GF180 MPW0 Serv) 6079 1 × 17t-5V 17,345 1,928 12,266 6% 9.00
Hash accelerator with clock multiplexing FF8E 1 × 17t-5V 12,962 1,441 8,276 5% 9.00
SPELL for GF180 B676 1 × 17t-5V 6,463 718 6,463 2% 9.00
ALU (paper) 66AD 1 × 17t-5V 5,960 662 5,960 1% 9.00
AS2650 105A 1 × 17t-5V 4,846 539 4,846 1% 9.00
Pyramiden Core (RV16U) F0DD 1 × 17t-5V 3,874 431 3,874 1% 9.00
PWM / Wishbone core 8BB5 1 × 17t-5V 2,471 275 2,471 1% 9.00
EduRISC16 E91D 1 × 17t-5V 2,034 226 2,004 <1% 9.00
Synchronous modulator 6E0D 1 × 17t-5V 1,558 173 1,558 <1% 9.00
Asynchronous modulator 9DDB 1 × 17t-5V 1,470 163 1,470 <1% 9.00
VGA Clock for GF180 0ABD 1 × 17t-5V 1,319 147 1,319 <1% 9.00
GreenRio (GF version) 9C7C 1 × 17t-5V 1,319 147 1,319 <1% 9.00
HyperRAM controller 1E9A 1 × 17t-5V 1,155 128 1,155 <1% 9.00

The projects

Every design on this run: an index, then grouped by what it does. Select one for its layout, its measurements and any die photographs.

Design index

All 40 public designs by their code, with the slot each took and how much logic it packs.

CodeDesignSlot What it isLogic cells Cells per mm²
0ABD VGA Clock for GF180 1 × 1 Retro computing 1,319 <1k
0AFF PDM driver and dice 1 × 1 Other digital 151 <1k
105A AS2650 1 × 1 Retro computing 4,846 1k
14FB SkullFET 1 × 1 Analog and mixed-signal 0
1E9A HyperRAM controller 1 × 1 Other digital 1,155 <1k
1F37 Beep Boop Traffic Light Controller 1 × 1 Other digital 154 <1k
4515 Tiny user project: inverter 1 × 1 Test structures 24 <1k
45D6 Ophelia eFPGA 1 × 1 FPGA 64,755 7k
472A ChipFlow example SoC with 3.3V cells 1 × 1 RISC-V and CPUs 727 <1k
4F9E MCU for solar power 1 × 1 Other digital 43 <1k
56C6 One-hot FABulous FPGA 1 × 1 FPGA 32,037 4k
58D2 TestFlipFlops 1 × 1 Other digital 28 <1k
5F3A Clock divider 1 × 1 Other digital 605 <1k
5FF2 74181 ALU 1 × 1 Other digital 77 <1k
6079 Serv tri-core (GF180 MPW0 Serv) 1 × 1 RISC-V and CPUs 17,345 2k
6106 FiveGuys 1 × 1 Other digital 105 <1k
66AD ALU (paper) 1 × 1 Other digital 5,960 1k
679D Rift2Fake 1 × 1 Other digital 308 <1k
6E0D Synchronous modulator 1 × 1 Other digital 1,558 <1k
6F7B Game of Life cell 1 × 1 Other digital 43 <1k
7F0C Universal gates 1 × 1 Other digital 182 <1k
8455 GF180 state machine 1 × 1 Other digital 88 <1k
8BB5 PWM / Wishbone core 1 × 1 Other digital 2,471 <1k
8D82 XLS bit population count 1 × 1 Multi-project die 48 <1k
9C7C GreenRio (GF version) 1 × 1 Other digital 1,319 <1k
9DDB Asynchronous modulator 1 × 1 Other digital 1,470 <1k
A006 5-bit multiplier 1 × 1 Other digital 163 <1k
A3A3 Signed 8-bit multiplier 1 × 1 Other digital 0
B22E TestAsyncTrimux 1 × 1 Other digital 89 <1k
B676 SPELL for GF180 1 × 1 Other digital 6,463 1k
BBE0 posoco2000 1 × 1 Other digital 115 <1k
C301 8x8 Bit Pattern Player 1 × 1 Other digital 201 <1k
C979 VerilogBoy 1 × 1 RISC-V and CPUs 27,097 3k
E91D EduRISC16 1 × 1 RISC-V and CPUs 2,034 <1k
EAD1 Reversible programmable logic IC 1 × 1 Other digital 44 <1k
EB3D Jacaranda-8 8-bit CPU 1 × 1 Retro computing 727 <1k
F0DD Pyramiden Core (RV16U) 1 × 1 RISC-V and CPUs 3,874 <1k
F330 16 x 4-bit dual port register file 1 × 1 Other digital 388 <1k
F4D8 OpenFASoC test structures in GF180MCU 1 × 1 Test structures 727 <1k
FF8E Hash accelerator with clock multiplexing 1 × 1 Accelerators 12,962 1k

Swipe the table sideways for more columns.

Designs that grew

These designs have been made more than once: first as a Tiny Tapeout tile or on a Google shuttle, then on wafer.space, or improved and made again on a later run. Start small, then come back bigger. The ringed marker is this run.

Tiny Tapeout tile Google shuttle wafer.space die this run
Each marker is one version of the design: hover over it or focus it for details, and select it to open its page. On a phone, the table below lists every one.

As a table
DesignMade onWhenVersionSize
FABulous FPGA Google GFMPW-0Dec 2022 56C6 One-hot FABulous FPGACaravel chip
wafer.space Run 1Dec 2025 MOLE FABulous FPGA1 × 1 slot
Tiny Tapeout IHP 26aMar 2026 Tiny FABulous FPGA8 × 4 tiles
Tiny Tapeout SKY 26aMay 2026 Tiny FABulous FPGA8 × 4 tiles
Tiny Tapeout SKY 26bMay 2026 Tiny FABulous FPGA8 × 4 tiles
Tiny Tapeout GF 0.3Jun 2026 Tiny FABulous FPGA; Tiny FABulous FPGA (3.3V version)4 × 4 tiles
wafer.space Run 2Jul 2026 MOLE FABulous FPGA1 × 1 slot
Tiny Tapeout SKY 26cSep 2026 Tiny FABulous FPGA8 × 4 tiles
Tiny Tapeout IHP 26bSep 2026 Tiny FABulous FPGA6 × 4 tiles
Avalon Semiconductors' retro-computing family Tiny Tapeout 2Dec 2022 Avalon Semiconductors '5401' 4-bit Microprocessor; Avalon Semiconductors 'TBB1143' Programmable Sound Generator1 × 1 tile
Google GFMPW-0Dec 2022 105A AS2650Caravel chip
Tiny Tapeout 3Apr 2023 Avalon Semiconductors '5401' 4-bit Microprocessor; Avalon Semiconductors 'TBB1143' Programmable Sound Generator1 × 1 tile
wafer.space Run 1Dec 2025 AS03 WS-Multi1 × 1 slot
wafer.space Run 2Jul 2026 AS04 Multi-Project die using 3.3V SCL1 × 1 slot
SERV multi-core RISC-V SoC Tiny Tapeout 2Dec 2022 serv - Serial RISC-V CPU1 × 1 tile
Google GFMPW-0Dec 2022 6079 Serv tri-core (GF180 MPW0 Serv)Caravel chip
Tiny Tapeout 3Apr 2023 serv - Serial RISC-V CPU1 × 1 tile
wafer.space Run 1Dec 2025 GD02 Racquet Half r1p0 (1/2 slot); GD03 Racquet r2p0 - 23 core SoC; GD04 Racquet Wide 1x0.5GD02 ½ × 1 slot, GD03 1 × 1 slot, GD04 1 × ½ slot
wafer.space Run 2Jul 2026 GD04 Racquet Wide 1x0.51 × ½ slot

Other digital 25

Show all 25 in Other digital

RISC-V and CPUs 5

Retro computing 3

FPGA 2

Test structures 2

Multi-project die 1

Analog and mixed-signal 1

Accelerators 1

The real silicon

Optical micrographs of the fabricated dies © 2025 John McMaster, CC BY 4.0

Microscope images © 2025 John McMaster, CC BY 4.0. Each project page links to a zoomable viewer of its die. The full-resolution images are online too.

Data and method

Everything on this page is measured from the public layout; the dates come from the Efabless GFMPW-0 shuttle page (archived) and the shuttle repositories.

  • Logic cells are standard cell instances, excluding fillers, taps, antenna diodes and tie cells.
  • Transistors are places where gate polysilicon crosses diffusion, not counting dummy fill.
  • Core area is the area of each die's user_project_wrapper cell.
  • Excluded from the measurements: pad ring (gf180mcu_*_io__*); caravel harness (caravel_core, caravel_power_routing, chip_io, caravel_logo, caravel_motto, copyright_block, open_source, user_id_textblock); seal ring (sealring). Die sizes are measured on the whole die, and bond pads on the whole die apart from its seal ring.
  • Bond pads are the pad openings in the pad ring, or in a second ring of pads inside it for chip-on-board bonding; probe pads are the others, next to the blocks they test. Letters and logos on the pad layer are not pads.
  • Theoretical maximum tiles one cell at its placement boundary with no routing.
  • Densest mm² is the best 1 mm square anywhere in each die, sliding in 0.1 mm steps and avoiding SRAM.