{"id":115121,"date":"2026-04-01T09:00:00","date_gmt":"2026-04-01T16:00:00","guid":{"rendered":"https:\/\/developer.nvidia.com\/blog\/?p=115121"},"modified":"2026-04-16T10:15:14","modified_gmt":"2026-04-16T17:15:14","slug":"cuda-tile-programming-now-available-for-basic","status":"publish","type":"post","link":"https:\/\/developer.nvidia.com\/blog\/cuda-tile-programming-now-available-for-basic\/","title":{"rendered":"CUDA Tile Programming Now Available for BASIC!"},"content":{"rendered":"\n

Note: <\/strong>CUDA Tile Programming in BASIC is an April Fools\u2019 joke, but it’s also real and actually works,\u00a0 demonstrating the flexibility of CUDA.<\/em><\/p>\n\n\n\n

CUDA 13.1<\/a> introduced CUDA Tile<\/a>, a next generation tile-based GPU programming paradigm designed to make fine-grained parallelism more accessible and flexible. One of its key strengths is language openness: any programming language can target CUDA Tile, enabling developers to bring tile-based GPU acceleration into a wide range of ecosystems.<\/p>\n\n\n\n

In response to overwhelming demand from seasoned developers everywhere, we\u2019re releasing cuTile BASIC<\/a> for GPUs, bringing CUDA Tile programming to this long-overlooked language.<\/p>\n\n\n\n

What is cuTile BASIC?<\/i><\/a><\/h2>\n\n\n\n

cuTile BASIC is an expression of the CUDA Tile programming model in BASIC, built on top of the CUDA Tile IR specification<\/a>. It enables you to write tile kernels in BASIC using a tile-based model, which is a natural fit for a programming language like BASIC which predates multi-threaded programming.<\/p>\n\n\n\n

cuTile BASIC is the perfect marriage of the power of GPUs with the anachronistic charm and syntactic simplicity of the BASIC programming language \u2013 an elegant language, from a more pixelated era. Manually numbering your lines of code never looked so good or ran so fast!<\/p>\n\n\n\n

Who is cuTile BASIC for?<\/i><\/a><\/h2>\n\n\n\n

BASIC is one of the oldest programming languages around and as such, is revered by a whole generation of developers who remember the sound of a 300 baud dial-up modem handshaking fondly. For many such developers, BASIC was their first introduction to computer programming.<\/p>\n\n\n\n

Now, developers with BASIC still burned into their brains can take legacy applications onto NVIDIA GPU-accelerated computing for the first time. This unlocks performance and functionality the BASIC programming language could never have previously imagined \u2013 allowing your Lunar Lander<\/a> to zip around the moon\u2019s surface faster than an Artemis mission.<\/p>\n\n\n\n

Relive the glory of writing games for your graphing calculator during math class while running on the world\u2019s most powerful GPUs.<\/p>\n\n\n\n

Get setup<\/i><\/a><\/h2>\n\n\n\n

First, install cuTile BASIC with PIP:<\/p>\n\n\n

\npip install git+https:\/\/github.com\/nvidia\/cuda-tile.git@basic-experimental\n<\/pre><\/div>\n\n\n
The full hardware and software requirements for running cuTile BASIC are listed at the end of this post (64k of RAM or more recommended). <\/p>\n\n\n\n
cuTile BASIC example<\/i><\/a><\/h2>\n\n\n\n
If you\u2019ve learned CUDA C++, you probably encountered the canonical vector addition kernel. A vector add kernel in CUDA C++ looks something like the following, which takes two vectors and adds them together elementwise to produce a third vector. <\/p>\n\n\n\n
This is one of the simplest CUDA kernels one can write:<\/p>\n\n\n
\n__global__ void vecAdd(float* A, float* B, float* C, int vectorLength)\n{\n \/* calculate my thread index *\/\n int workIndex = threadIdx.x + blockIdx.x*blockDim.x;\n\n if(workIndex < vectorLength)\n {\n  \/* perform the vector addition *\/\n  C[workIndex] = A[workIndex] + B[workIndex];\n }\n}\n<\/pre><\/div>\n\n\n
In this kernel, each thread\u2019s work is explicitly specified, and the programmer, when launching this kernel, will specify the number of blocks and threads to be launched.<\/p>\n\n\n\n
Now let\u2019s look at the equivalent code written in cuTile BASIC. We don\u2019t need to specify what each thread does. We only have to break the data into tiles and specify what mathematical operations should happen to these tiles. Everything else is handled for us.<\/p>\n\n\n\n
The cuTile BASIC vector add kernel is shown below:<\/p>\n\n\n
\n10 REM Vector Add: C = A + B\n20 INPUT N, A(), B()\n30 DIM A(N), B(N), C(N)\n40 TILE A(128), B(128), C(128)\n50 LET C(BID) = A(BID) + B(BID)\n60 OUTPUT C\n70 END\n<\/pre><\/div>\n\n\n
This example is very basic and uses standard BASIC with three additional things to point out:<\/p>\n\n\n\n