You don't need a board to start FPGA design

The most common question from people getting into FPGAs is "which board should I buy?" It feels like the natural first step: the chip is the point, so surely you need the hardware in hand. The surprising and genuinely freeing answer is that you don't need a board at all to start. You can learn most of what matters, and get quite good, before you ever buy silicon.

This post is inspired by a thread from Adam Taylor (@ATaylorFPGA), a chartered engineer who has taught a lot of people this craft. His point lands hard because it is counter-intuitive: the fastest way in is not a shopping cart, it is a simulator.

Why simulation is the better place to begin

It seems backwards until you have felt it, and then it is obvious.

You see everything. On a board, a signal deep inside your design is invisible unless you route it to a pin or wire up a logic analyzer. In simulation, every wire and every register is right there in the waveform, at every moment in time. You learn far faster when you can watch the logic think.
The loop is seconds, not minutes. Change a line, re-simulate, read the result. No synthesis, no place-and-route, no bitstream, no reprogramming. The tight feedback loop is where learning actually happens.
You are learning the real skill. FPGA design is not about the board. It is about describing hardware in an HDL and reasoning about concurrency, timing, and state. That skill is identical whether the design ends up on a $30 dev board or a datacenter accelerator. Start with the part that transfers.
Nothing to break, nothing to blame. No wiring mistakes, no power supply, no "is it the board or my code?" A simulator removes every variable except the one you are trying to learn.

What you can do today with zero hardware

Quite a lot, as it turns out. With nothing but a browser you can:

Learn an HDL. Verilog or VHDL, the two languages of the field. Our free course walks from logic gates to a working ALU, every lesson in both languages with a real simulated waveform. If you think in Python, MyHDL lets you describe hardware in a language you already know.
Simulate and watch the waveform. The Verilog playground compiles and runs your code on our server and draws the trace, no installs, no toolchain, no account. The MyHDL playground does the same for Python hardware.
Lint it. Catch the classic mistakes (a latch you did not mean to infer, a width mismatch) before they ever reach a board, with the built-in Verilator lint.
Synthesize for a resource estimate. See roughly how many LUTs and flip-flops your design costs, so you build intuition for size and speed without owning the chip.
Prove you can actually do it. Our challenges hand you a spec and a hidden testbench and grade your module the moment it passes. That is the real test of learning, and it needs no hardware.

Your first hour, concretely

If you are starting from zero, here is a path that costs nothing:

Read the first two lessons (logic gates, then multiplexers) and actually open each one in the playground and change something.
Write a 4-bit counter from scratch in the playground and watch it count in the waveform.
Break it on purpose: remove the reset, mismatch a width, and see what the linter says. Learning what "wrong" looks like is half the battle.
Attempt the first challenge. Passing a locked testbench is a genuinely different feeling from "it looked right."

Every term you do not recognize along the way is one click away in the glossary.

So when do you actually want a board?

Eventually, and it is worth it. Hardware teaches the things simulation cannot: real timing closure against a real clock, I/O standards, getting a design to fit and run at speed, and the simple joy of blinking a physical LED with logic you wrote. Adam Taylor's own work, including developing FPGAs on a Raspberry Pi 400, shows how little you need to spend even when you do move to hardware.

The point is only about order. Buy the board when you have a design you are excited to see run, not as the price of admission. When that day comes, our dev board picker helps you choose one that fits your goals and budget.

Start now, in the tab you already have open

The barrier to FPGA design was never the hardware. It was the belief that you needed it first. You do not. Open the playground, write your first module, and watch it work. The board can wait.

Inspired by Adam Taylor's thread on getting started in FPGA. He writes and teaches prolifically at Adiuvo and on Hackster; well worth following.