doc: updated the list of supported hardware

README.binary.in

@@ -4,7 +4,7 @@ oneAPI Deep Neural Network Library (oneDNN)
 oneAPI Deep Neural Network Library (oneDNN) is an
 open-source performance library for deep learning applications. The library
 includes basic building blocks for neural networks optimized
-for Intel Architecture Processors, Intel Processor Graphics, and 
+for Intel(R) Architecture Processors, Intel Processor Graphics, and 
 Xe Architecture graphics.
 
 This package contains oneDNN v@PROJECT_VERSION@ (@DNNL_VERSION_HASH@).
@@ -27,11 +27,11 @@ System Requirements
 oneDNN supports systems based on Intel 64 or AMD64 architectures.
 
 The library is optimized for the following CPUs:
-* Intel Atom processor with Intel SSE4.1 support
-* 4th, 5th, 6th, 7th, and 8th generation Intel(R) Core(TM) processor
-* Intel(R) Xeon(R) processor E3, E5, and E7 family (formerly Sandy Bridge,
+* Intel Atom processors (at least Intel SSE4.1 support is required)
+* Intel Core(TM) processors (at least Intel SSE4.1 support is required)
+* Intel Xeon(R) processor E3, E5, and E7 family (formerly Sandy Bridge,
   Ivy Bridge, Haswell, and Broadwell)
-* Intel(R) Xeon Phi(TM) processor (formerly Knights Landing and Knights Mill)
+* Intel Xeon Phi(TM) processor (formerly Knights Landing and Knights Mill)
 * Intel Xeon Scalable processor (formerly Skylake, Cascade Lake, and Cooper
   Lake)
 * future Intel Xeon Scalable processor (code name Sapphire Rapids)
@@ -44,10 +44,9 @@ them. See CPU dispatcher control (https://oneapi-src.github.io/oneDNN/dev_guide_
 for more details.
 
 The library is optimized for the following GPUs:
-* Intel HD Graphics
-* Intel UHD Graphics
-* Intel Iris Plus Graphics
-* Xe Architecture graphics (code named DG1 and Tiger Lake)
+* Intel Processor Graphics based on Gen9, Gen9.5 and Gen11, and Gen12 architectures
+* Intel Iris(R) Xe graphics (formerly DG1)
+* future Intel Arc(TM) graphics (code name Alchemist and DG2)
 
 ## Linux
 

README.md

@@ -10,10 +10,10 @@ oneAPI Deep Neural Network Library (oneDNN)
 oneAPI Deep Neural Network Library (oneDNN) is an open-source cross-platform
 performance library of basic building blocks for deep learning applications.
 oneDNN is part of [oneAPI](https://oneapi.io).
-The library is optimized for Intel Architecture Processors, Intel Processor
+The library is optimized for Intel(R) Architecture Processors, Intel Processor
 Graphics and Xe Architecture graphics. oneDNN has experimental support
 for the following architectures: Arm\* 64-bit Architecture (AArch64),
-NVIDIA\* GPU, OpenPOWER\* Power ISA (PPC64), IBMz\* (s390x). 
+NVIDIA\* GPU, OpenPOWER\* Power ISA (PPC64), IBMz\* (s390x), and RISC-V.
 
 oneDNN is intended for deep learning applications and framework
 developers interested in improving application performance
@@ -71,11 +71,11 @@ oneDNN supports platforms based on the following architectures:
 > and RISC-V (RV64) support is **experimental** with limited testing validation.
 
 The library is optimized for the following CPUs:
-* Intel Atom processor with Intel SSE4.1 support
-* 4th, 5th, 6th, 7th, and 8th generation Intel(R) Core(TM) processor
-* Intel(R) Xeon(R) processor E3, E5, and E7 family (formerly Sandy Bridge,
+* Intel Atom(R) processors (at least Intel SSE4.1 support is required)
+* Intel Core(TM) processors (at least Intel SSE4.1 support is required)
+* Intel Xeon(R) processor E3, E5, and E7 family (formerly Sandy Bridge,
   Ivy Bridge, Haswell, and Broadwell)
-* Intel(R) Xeon Phi(TM) processor (formerly Knights Landing and Knights Mill)
+* Intel Xeon Phi(TM) processor (formerly Knights Landing and Knights Mill)
 * Intel Xeon Scalable processor (formerly Skylake, Cascade Lake, and Cooper
   Lake)
 * future Intel Xeon Scalable processor (code name Sapphire Rapids)
@@ -102,9 +102,10 @@ compatible with Compute Library versions 21.08 or later.
 > [linking guide](https://oneapi-src.github.io/oneDNN/dev_guide_link.html)
 > for more details.
 
-The library is optimized for the following GPU architectures:
-* Intel Processor Graphics Gen9, Gen9.5 and Gen11 architectures
-* Xe architecture
+The library is optimized for the following GPUs:
+* Intel Processor Graphics based on Gen9, Gen9.5 and Gen11, and Gen12 architectures
+* Intel Iris(R) Xe graphics (formerly DG1)
+* future Intel Arc(TM) graphics (code name Alchemist and DG2)
 
 ## Requirements for Building from Source
 

doc/programming_model/data_types.md

@@ -131,16 +131,18 @@ types that oneDNN recognizes.
   zero.
 
 ### Intel(R) Processor Graphics and Xe Architecture graphics
-
-Intel Processor Graphics provides hardware acceleration for fp32 and fp16
-arithmetic. Xe Architecture graphics additionally provides
-acceleration for int8 arithmetic (both signed and unsigned). Implementations
-for the bf16 data type are functional only and do not currently provide
-performance benefits.
-
-| Data type | Support level
+oneDNN performance optimizations for Intel Processor graphics and
+Xe Architecture graphics are specialized based on device microarchitecture (uArch).
+The following uArchs have specialized optimizations in the library:
+* GEN9 (also covers GEN11)
+* Xe-LP (previoulsy known as GEN12LP)
+* Xe-HP
+
+The following table indicates the minimal supported uArch for each of the data
+types that oneDNN recognizes.
+| Data type | Minimal supported uArch
 | :---      | :---
-| f32       | optimized
-| bf16      | functional only
-| f16       | optimized
-| s8, u8    | optimized for Xe Architecture graphics (code named DG1 and Tiger Lake)
+| f32       | GEN9
+| s8, u8    | Xe-LP
+| bf16      | Xe-HP
+| f16       | GEN9