At CES 2025 AMD has added new processors to its Zen 5 based Ryzen AI series with the announcement of a comprehensive line-up of Ryzen AI processors – AMD Ryzen AI Max, Ryzen AI 300 Series and Ryzen 200 Series processors. Along with the Pro versions – Ryzen AI Max PRO, Ryzen AI 300 PRO and Ryzen AI 200 PRO.
Yeah, we also said in our mind – Are you serious AMD with this naming scheme? This launch builds upon AMD’s previous announcement in October 2024, which saw launch of Ryzen AI Pro 300 – AI 9 HX PRO 375, AI 9 HX PRO 370 and AI 7 PRO 360.
Quick Preview
- Three New Ryzen AI Series: Max (and Max PRO), 300 (and 300 PRO), and 200 (and 200 PRO),
- Zen 5″ and “Zen 4” Architectures: The Max and 300 series are based on the latest “Zen 5” architecture, while the 200 series uses “Zen 4” architecture.
- Core Counts up to 16: Offering up to 16 cores (and 32 threads) in the high-end Max series.
- Building on Previous Announcements: This launch builds upon the October 2024 reveal of the Ryzen AI PRO 300 series, now expanding it to a wider range of consumers and businesses.
- Ryzen PRO Series processors feature enterprise-grade security and manageability tools designed to help secure the modern enterprise and streamline IT operations.
Ryzen AI Max and Max PRO: The Flagship
The Ryzen AI Max series represents the pinnacle of AMD’s mobile AI offerings, designed for users who demand uncompromising performance on the go. The addition of the Max PRO variants extends this power to professional workstations.
Models include the AI Max+ 395, AI Max 390, AI Max 385, and their PRO counterparts.
Specifications of AMD Ryzen AI Max and Max Pro processors:
- Up to 16 “Zen 5” CPU cores for exceptional multithreaded performance.
- Up to 40 AMD RDNA 3.5 graphics compute units, providing a significant boost in graphics capabilities for gaming, content creation, and professional visualization.
- An XDNA 2 NPU with up to 50 TOPS of dedicated AI processing power, crucial for accelerating AI-driven workflows.
- Support for up to 128GB of unified memory, with up to 96GB available for graphics, enabling seamless multitasking and handling of large AI models, especially important for professional applications.
Model | Cores / Threads | Boost2 / Base Frequency | Total Cache | Graphics Model AMD | cTDP | NPU TOPS | Graphics Cores |
AMD Ryzen™ AI Max+ 395 | 16C/32T | Up to 5.1 /3.0 GHz | 80MB | AMD Radeon™ 8060S Graphics | 45-120W | 50 | 40 |
AMD Ryzen™ AI Max 390 | 12C/24T | Up to 5.0 / 3.2 GHz | 76MB | AMD Radeon™ 8050S Graphics | 45-120W | 50 | 32 |
AMD Ryzen™ AI Max 385 | 8C/16T | Up to 5.0 / 3.6 GHz | 40MB | AMD Radeon™ 8050S Graphics | 45-120W | 50 | 32 |
AMD Ryzen™ AI Max+ PRO 395 | 16C/32T | Up to 5.1 / 3.0 GHz | 80MB | AMD Radeon™ 8060S Graphics | 45-120W | 50 | 40 |
AMD Ryzen™ AI Max PRO 390 | 12C/24T | Up to 5.0 / 3.2 GHz | 76MB | AMD Radeon™ 8050S Graphics | 45-120W | 50 | 32 |
AMD Ryzen™ AI Max PRO 385 | 8C/16T | Up to 5.0 / 3.6 GHz | 40MB | AMD Radeon™ 8050S Graphics | 45-120W | 50 | 32 |
AMD Ryzen™ AI Max PRO 380 | 6C/12T | Up to 4.9 / 3.6 GHz | 22MB | AMD Radeon™ 8040S Graphics | 45-120W | 50 | 16 |
Availability
Systems powered by Ryzen AI Max and Ryzen AI Max PRO Series Processors are expected to be available starting in Q1 2025.
The Max PRO variants specifically cater to business needs with AMD PRO Technologies, including enhanced security features, robust manageability tools for IT departments, and features like Cloud Bare Metal Recovery. This makes them ideal for professionals working with large engineering models, architectural designs, and complex AI-accelerated workloads.
Ryzen AI 300 and 300 PRO: Expanding the AI Ecosystem for Consumers and Businesses
AMD initially previewed the Ryzen AI PRO 300 series in June 2024, highlighting its focus on business productivity and integration with Microsoft Copilot+ features. This announcement focused on the higher-end SKUs like the Ryzen AI 9 HX PRO 375, emphasizing its leading AI performance and suitability for demanding business tasks.
The CES 2025 announcement significantly expands this series, introducing additional models, including the Ryzen AI 7 350 and Ryzen AI 5 340, making these powerful AI capabilities available to a wider consumer audience. The PRO variants continue to offer the crucial AMD PRO Technologies.
Key features of the Ryzen AI 300 series include:
- Up to 8 “Zen 5” CPU cores.
- The latest RDNA 3.5 graphics architecture.
- An industry-leading NPU powered by AMD XDNA 2 technology, delivering up to five times better performance than the first-generation NPU.
- Support for Microsoft Copilot+ experiences, such as live captioning and language translation, benefiting both consumer and business users.
Model | Cores / Threads | Boost5 / Base Frequency | Total Cache | Graphics Model AMD | cTDP | NPU TOPS |
AMD Ryzen™ AI 7 350 | 8C/16T | Up to 5.0 / 2.0 GHz | 24 MB | AMD Radeon™ 860M Graphics | 15-54W | 50 |
AMD Ryzen™ AI 5 340 | 6C/12T | Up to 4.8 / 2.0 GHz | 22 MB | AMD Radeon™ 840M Graphics | 15-54W | 50 |
AMD Ryzen™ AI 7 PRO 350 | 8C/16T | Up to 5.0 / 2.0 GHz | 24 MB | AMD Radeon™ 860M Graphics | 15-54W | 50 |
AMD Ryzen™ AI 5 PRO 340 | 6C/12T | Up to 4.8 / 2.0 GHz | 22 MB | AMD Radeon™ 840M Graphics | 15-54W | 50 |
Availability
Systems powered by the new Ryzen AI 300 processors are expected to be available starting in Q1 2025.
The Ryzen AI 300 PRO series is specifically tailored for commercial use, offering enhanced security and manageability features through AMD PRO Technologies. This builds upon the October announcement by broadening the availability of these features across more SKUs.
Ryzen 200 and 200 PRO: Democratizing AI for Mainstream Users and Businesses with Essential Needs
AMD is also bringing AI capabilities to more mainstream devices with the Ryzen 200 series, and again offering PRO variants for business users with more basic needs but still requiring the security and management features. These processors leverage the proven “Zen 4” architecture and the FP8 platform to offer a compelling blend of performance and efficiency at more accessible price points. While their NPU TOPS are lower, they still provide a significant boost for AI-accelerated tasks in everyday applications.
Key features of the Ryzen 200 series include:
- Up to 8 CPU cores and 16 threads.
- AMD RDNA 3 graphics.
- Up to 16 NPU TOPS of AI processing power.
Model | Cores / Threads | Boost6 / Base Frequency | Total Cache | Graphics Model AMD | cTDP | NPU TOPS |
AMD Ryzen™ 9 270 | 8C/16T | Up to 5.2 / 4.0 GHz | 24MB | AMD Radeon™ 780M Graphics | 35-54W | 16 |
AMD Ryzen™ 7 260 | 8C/16T | Up to 5.1 / 3.8 GHz | 24MB | AMD Radeon™ 780M Graphics | 35-54W | 16 |
AMD Ryzen™ 7 250 | 8C/16T | Up to 5.1 / 3.3 GHz | 24MB | AMD Radeon™ 780M Graphics | 15-30W | 16 |
AMD Ryzen™ 5 240 | 6C/12T | Up to 5.0 / 4.3 GHz | 22MB | AMD Radeon™ 760M Graphics | 35-54W | 16 |
AMD Ryzen™ 5 230 | 6C/12T | Up to 4.9 / 3.5 GHz | 22MB | AMD Radeon™ 760M Graphics | 15-30W | 16 |
AMD Ryzen™ 5 220 | 6C/12T | Up to 4.9 / 3.2 GHz | 22MB | AMD Radeon™ 740M Graphics | 15-30W | N/A |
AMD Ryzen™ 3 210 | 4C/8T | Up to 4.7 / 3.0 GHz | 12MB | AMD Radeon™ 740M Graphics | 15-30W | N/A |
AMD Ryzen™ 7 PRO 250 | 8C/16T | Up to 5.1 / 3.3 GHz | 24 MB | AMD Radeon™ 780M graphics | 15-30W | 16 |
AMD Ryzen™ 5 PRO 230 | 6C/12T | Up to 4.9 / 3.5 GHz | 22 MB | AMD Radeon™ 760M graphics | 15-30W | 16 |
AMD Ryzen™ 5 PRO 220 | 6C/12T | Up to 4.9 / 3.2 GHz | 22 MB | AMD Radeon™ 740M graphics | 15-30W | N/A |
AMD Ryzen™ 3 PRO 210 | 4C/8T | Up to 4.7 / 3 GHz | 12 MB | AMD Radeon™ 740M graphics | 15-30W | N/A |
Availability
Systems powered by Ryzen 200 and Ryzen 200 PRO Series processors are expected to be available starting in Q2 2025.
The Ryzen 200 PRO series provides essential business features via AMD PRO Technologies, making it suitable for deployments where robust security and IT manageability are required, but high-end AI performance is not the primary concern.
The Ryzen 200 series targets the mainstream market, bringing AI capabilities to more affordable laptops. Using the older but still plenty powerful “Zen 4” architecture, these processors offer a balance of performance and efficiency, with an integrated NPU for accelerating AI tasks in everyday applications.
Clearing out the confusion : Complete list of AMD Ryzen AI processors
Now with all the AMD Ryzen AI processors. There is a lot of confusion. So, we have compiled this table of all the models.
Processor Series | Model | Cores/Threads | Boost/Base Frequency (GHz) | Total Cache | Graphics Model | cTDP (W) | NPU TOPS | Announcement Date |
Ryzen AI Max | AI Max+ 395 | 16C/32T | Up to 5.1 / 3.0 | 80MB | Radeon 8060S | 45-120 | 50 | Jan 2025 (CES) |
AI Max 390 | 12C/24T | Up to 5.0 / 3.2 | 76MB | Radeon 8050S | 45-120 | 50 | Jan 2025 (CES) | |
AI Max 385 | 8C/16T | Up to 5.0 / 3.6 | 40MB | Radeon 8050S | 45-120 | 50 | Jan 2025 (CES) | |
Ryzen AI Max PRO | AI Max+ PRO 395 | 16C/32T | Up to 5.1 / 3.0 | 80MB | Radeon 8060S | 45-120 | 50 | Jan 2025 (CES) |
AI Max PRO 390 | 12C/24T | Up to 5.0 / 3.2 | 76MB | Radeon 8050S | 45-120 | 50 | Jan 2025 (CES) | |
AI Max PRO 385 | 8C/16T | Up to 5.0 / 3.6 | 40MB | Radeon 8050S | 45-120 | 50 | Jan 2025 (CES) | |
AI Max PRO 380 | 6C/12T | Up to 4.9 / 3.6 | 22MB | Radeon 8040S | 45-120 | 50 | Jan 2025 (CES) | |
Ryzen AI 300 | AI 7 350 | 8C/16T | Up to 5.0 / 2.0 | 24MB | Radeon 860M | 15-54 | 50 | Jan 2025 (CES) |
AI 5 340 | 6C/12T | Up to 4.8 / 2.0 | 22MB | Radeon 840M | 15-54 | 50 | Jan 2025 (CES) | |
Ryzen AI 300 PRO | AI 9 HX PRO 375 | 12C/24T | Up to 5.1 / 2.0 | 36MB | Radeon 890M | 15-54 | Up to 55 | Oct 2024 |
AI 9 HX PRO 370 | 12C/24T | Up to 5.1 / 2.0 | 36MB | Radeon 890M | 15-54 | Up to 50 | Oct 2024 | |
AI 7 PRO 360 | 8C/16T | Up to 5.0 / 2.0 | 24MB | Radeon 880M | 15-54 | Up to 50 | Oct 2024 | |
AI 7 PRO 350 | 8C/16T | Up to 5.0 / 2.0 | 24MB | Radeon 860M | 15-54 | 50 | Jan 2025 (CES) | |
AI 5 PRO 340 | 6C/12T | Up to 4.8 / 2.0 | 22MB | Radeon 840M | 15-54 | 50 | Jan 2025 (CES) | |
Ryzen 200 | 9 270 | 8C/16T | Up to 5.2 / 4.0 | 24MB | Radeon 780M | 35-54 | 16 | Jan 2025 (CES) |
7 260 | 8C/16T | Up to 5.1 / 3.8 | 24MB | Radeon 780M | 35-54 | 16 | Jan 2025 (CES) | |
7 250 | 8C/16T | Up to 5.1 / 3.3 | 24MB | Radeon 780M | 15-30 | 16 | Jan 2025 (CES) | |
5 240 | 6C/12T | Up to 5.0 / 4.3 | 22MB | Radeon 760M | 35-54 | 16 | Jan 2025 (CES) | |
5 230 | 6C/12T | Up to 4.9 / 3.5 | 22MB | Radeon 760M | 15-30 | 16 | Jan 2025 (CES) | |
5 220 | 6C/12T | Up to 4.9 / 3.2 | 22MB | Radeon 740M | 15-30 | N/A | Jan 2025 (CES) | |
3 210 | 4C/8T | Up to 4.7 / 3.0 | 12MB | Radeon 740M | 15-30 | N/A | Jan 2025 (CES) | |
Ryzen 200 PRO | 7 PRO 250 | 8C/16T | Up to 5.1 / 3.3 | 24MB | Radeon 780M | 15-30 | 16 | Jan 2025 (CES) |
5 PRO 230 | 6C/12T | Up to 4.9 / 3.5 | 22MB | Radeon 760M | 15-30 | 16 | Jan 2025 (CES) | |
5 PRO 220 | 6C/12T | Up to 4.9 / 3.2 | 22MB | Radeon 740M | 15-30 | N/A | Jan 2025 (CES) | |
3 PRO 210 | 4C/8T | Up to 4.7 / 3.0 | 12MB | Radeon 740M | 15-30 | N/A | Jan 2025 (CES) |
AMDs AI push: a decent strategy, but the naming is confusing
When most people think of AI, they think of NVIDIA. NVIDIA has undeniably dominated the AI market, particularly in data centers and high-performance computing, thanks to its powerful GPUs and CUDA software platform. But AMD is taking a different approach. While AMD also offers AI datacenter AI, AMD is focusing on bringing AI acceleration directly to the PC.
The Naming Problem: A Recipe for Confusion
AMD’s naming scheme is overly complex, creating unnecessary confusion:
- “AI” Overload: The constant use of “AI” (“Ryzen AI,” “Ryzen AI Max”) blurs the line between the general AI market and the specific on-device AI acceleration these chips provide.
- Layered Designations: Combining “Max” (high-performance consumer), “Pro” (commercial), and “HX” (higher TDP) with numerical designations creates unwieldy names.
- Numerical Chaos: The numerical designations (395, 390, 375, etc.) lack a clear, intuitive pattern, making comparisons difficult.
This results in names like “Ryzen 9 HX PRO 375” or “Ryzen 7 AI Pro 350″—a mouthful that doesn’t effectively communicate key differences.
While Intel has integrated AI acceleration into its latest Core Ultra processors, AMD’s dedicated NPU approach offers a potentially more powerful and efficient solution for handling AI workloads.
Leave a Reply
You must be logged in to post a comment.