This is an automated email from the ASF dual-hosted git repository.
tqchen pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/tvm-site.git
The following commit(s) were added to refs/heads/main by this push:
new dfefd9df994 Refresh the front page and menu around TVM TIRx and TVM FFI
dfefd9df994 is described below
commit dfefd9df9944c740d5a4437cf2613a93533823ea
Author: tqchen <[email protected]>
AuthorDate: Wed Nov 11 20:54:08 2026 +0000
Refresh the front page and menu around TVM TIRx and TVM FFI
The hero describes TVM as modular, minimal, and stable components for
building ML compilers in the agentic era. The front page lists TVM TIRx and TVM
FFI as its projects and drops the old figure, feature tiles, and link blocks.
The menu leads with TVM TIRx and TVM FFI.
---
_data/menus.yml | 18 ++++++------
_includes/header_index.html | 5 +++-
css/custom.scss | 40 ++++++++++++++++++++++++++
index.md | 69 ++++++---------------------------------------
4 files changed, 62 insertions(+), 70 deletions(-)
diff --git a/_data/menus.yml b/_data/menus.yml
index 648050279e9..e573bed9763 100644
--- a/_data/menus.yml
+++ b/_data/menus.yml
@@ -1,14 +1,14 @@
-- name: Community
- link: /community
-- name: Download
- link: /download
-- name: Docs
- link: https://tvm.apache.org/docs/
-- name: Posts
- link: /posts
- name: TVM TIRx
link: https://github.com/apache/tvm-tirx/
- name: TVM FFI
link: https://github.com/apache/tvm-ffi/
-- name: Github
+- name: Download
+ link: /download
+- name: Posts
+ link: /posts
+- name: Community
+ link: /community
+- name: Docs
+ link: https://tvm.apache.org/docs/
+- name: GitHub
link: https://github.com/apache/tvm/
diff --git a/_includes/header_index.html b/_includes/header_index.html
index f411db5b0ba..b9d5a281ca8 100644
--- a/_includes/header_index.html
+++ b/_includes/header_index.html
@@ -11,7 +11,10 @@
</div>
<div class="rightAlignDetails">
<p class="subHeadingTwo">
- Apache TVM is a machine learning compilation framework,
following the principle of Python-first development and universal deployment.
It takes in pre-trained machine learning models, compiles and generates
deployable modules that can be embedded and run everywhere.
+ Apache TVM provides modular, minimal, and stable components for
building machine
+ learning compilers in the agentic era: an open, stable ABI that
connects compilers,
+ runtimes, and languages, and a compiler foundation for building
ML compilers, kernel
+ DSLs, and the tools that analyze and verify their programs.
</p>
</div>
</div>
diff --git a/css/custom.scss b/css/custom.scss
index 22e5ac01e4c..6e90099ddfb 100644
--- a/css/custom.scss
+++ b/css/custom.scss
@@ -731,6 +731,46 @@ p {
}
}
}
+// The about section without its figure: one full-width text column.
+.aboutSec.aboutTextOnly {
+ padding: 90px 0;
+ @media #{$screen-sm} {
+ padding: 0;
+ }
+ &::after {
+ display: none;
+ }
+ .aboutInner .aboutDetailsCol {
+ flex: 1 1 auto;
+ max-width: 860px;
+ padding-left: 0;
+ @media #{$screen-sm} {
+ padding: 58px 30px 0;
+ }
+ h4 {
+ font-size: 32px;
+ line-height: 40px;
+ }
+ a {
+ color: $darkBule;
+ text-decoration: underline;
+ }
+ // Projects as plain entries with an accent bar, not numbered.
+ ul li {
+ padding-left: 20px;
+ border-left: 3px solid $darkBule;
+ @media #{$screen-md} {
+ padding-left: 20px;
+ }
+ @media #{$screen-sm} {
+ padding-left: 16px;
+ }
+ &::after {
+ content: none;
+ }
+ }
+ }
+}
//about css end
// keysection css start
.keyfeatures {
diff --git a/index.md b/index.md
index 1f9da6e77fc..d726b93246b 100644
--- a/index.md
+++ b/index.md
@@ -5,69 +5,18 @@ title: "Apache TVM"
{::options parse_block_html="true" /}
-<section class="aboutSec">
+<section class="aboutSec aboutTextOnly">
<div class="container">
{:.aboutInner}
-* {:.aboutImgCol}
- {:.desktopImg}
- {:.responsiveImg}
* {:.aboutDetailsCol}
-#### About Apache TVM
-The vision of the Apache TVM Project is to host a diverse community of experts
and practitioners
-in machine learning, compilers, and systems architecture to build an
accessible, flexible, and
-efficient open-source framework that optimizes current and emerging machine
learning models for
-any hardware platform. TVM provides the following main features:
-
- * Python-first development that enables quick customization of machine
learning compiler pipelines.
- * Universal deployment to bring models into minimum deployable modules.
-</div>
-</section>
-
-<section class="keyfeatures">
-<div class="key-section-container">
-{:.key-title-text}
-## Key Features & Capabilities
-
-* {:.key-block}
-
-### Performance
-
- {:.mb-3}
- Compilation and minimal runtimes commonly unlock ML workloads on existing
hardware.
-
-* {:.key-block}
-
-### Run Everywhere
- From data center GPUs to edge environments.
-
-
-* {:.key-block}
-
-### Flexibility
- Flexible design that enables easy customization of compilation pipelines.
-
-* {:.key-block}
-
-### Ease of Use
- Easy to use python first compiler API that brings universal deployment.
-
-</div>
-
-</section>
-
-
-<section class="docSec">
-<div class="container">
-
-* {:.doc-link-block}
-### Docs
-[Written with care <br/> & love for you.](https://tvm.apache.org/docs/)
-
-
-* {:.doc-link-block}
-### Community
-[Join the TVM <br/> community](/community)
-
+#### Projects
+
+ * **[TVM TIRx](https://github.com/apache/tvm-tirx/)**: the ML compiler
foundation. A minimal,
+ stable core IR with an open operator system, utilities for symbolic
analysis, structural
+ traversal and mapping, and code generation for CUDA and host LLVM
targets.
+ * **[TVM FFI](https://github.com/apache/tvm-ffi/)**: an open ABI and FFI
for ML systems. It
+ lets compiler objects, functions, and tensors move across Python, C++,
Rust, and
+ frameworks such as PyTorch.
</div>
</section>