AMD recently introduced its new ROCm version 6.1.3. This version brings significant benefits for developers and companies in the field of AI workloads. Those working in the field of artificial intelligence can benefit in particular, as the new ROCm version offers numerous improvements and new functions.

The new release focuses on improving compatibility, accessibility and scalability. One highlight is multi-GPU support, which enables the creation of scalable AI desktops for multi-serving and multi-user solutions. This support includes up to four qualified Radeon RX-Series or Radeon PRO GPUs running simultaneously. Users now have the ability to utilize configurations with data parallelism, where each GPU acts independently and performs different calculations. This leads to more versatile and efficient AI applications supported by AMD ROCm software and Radeon GPUs.

In addition, the new ROCm version now supports TensorFlow, which further facilitates the development of AI applications and offers more options for developers. Another highlight is the introduction of the AMD Radeon PRO W7900 dual slot card, which offers 192 AI accelerators and 48GB of GDDR6 memory in a compact form factor, enabling higher density at the system level. This graphics card expands AMD’s offering for client-based AI development.

For users who want to work on Linux but only have a Windows system, ROCm now offers support for the Windows Subsystem for Linux, also known as WSL 2. This makes it possible to run Linux-based AI tools on a Windows system without the need for a dedicated Linux system or dual-boot setup.
Andrej Zdravkovic, Senior Vice President at AMD, explains that the new ROCm release extends functional parity from the data center to desktops, making AI research much easier.
“The new AMD ROCm release extends functional parity from data center to desktops, enabling AI research and development on readily available and accessible platforms”
Source: AMD
15 Antworten
Kommentar
Lade neue Kommentare
Urgestein
Veteran
Veteran
Mitglied
Urgestein
Mitglied
Mitglied
Urgestein
Mitglied
Urgestein
Mitglied
Urgestein
Mitglied
Veteran
Urgestein
Alle Kommentare lesen unter igor´sLAB Community →