Listen to this Post
AMD has once again made waves in the AI computing world with its latest APU, the Ryzen AI Max+ 395. Marketed as the “most powerful x86 APU” available, it boasts significant advantages over its Intel counterpart, the Core Ultra 7 258V. With official benchmarks showcasing AMD’s dominance in AI workloads, the question remains: Is this truly a game-changer, or just a case of selective comparisons?
The Ryzen AI Max+ 395 comes equipped with 16 cores, 32 threads, a powerful XDNA 2 NPU with 50+ AI TOPS, and Radeon 8060S integrated graphics. This APU has been designed to excel in AI-driven tasks, particularly in large language model (LLM) processing. However, some industry experts are questioning the fairness of AMD’s comparisons, given the hardware and memory disparities between tested devices.
This article explores AMD’s benchmark results, analyzes the performance difference, and provides an in-depth look at whether Intel has any real competition in this space.
AMD Ryzen AI Max+ 395: Benchmark Results and Performance Insights
AMD’s Official Benchmarks
AMD tested the Ryzen AI Max+ 395 against the Intel Core Ultra 7 258V using LM Studio (version 0.3.11), running multiple LLM models, including DeepSeek R1, Phi 4 Mini Instruct, and Llama 3.2. The benchmarks were measured using two key AI performance metrics:
– Tokens per second (throughput efficiency)
- Time to first token (latency and processing speed)
In these tests, AMD’s Ryzen AI Max+ 395 delivered at least twice the tokens per second compared to the Intel Core Ultra 7 258V. The biggest performance gap appeared in DeepSeek R1 Distill Qwen 14b, where AMD’s processor was 12.2x faster than Intel’s. Similarly, Phi 4 14b showed an 11.3x improvement in speed with AMD’s chip.
While the Ryzen AI Max+ 395 dominated in most scenarios, its lead varied across different AI models. For example, performance in Llama 3.2 ranged from 4x to 9x faster than the Intel Core Ultra 7 258V.
Comparing Hardware: AMD vs. Intel
A major factor contributing to this performance difference is hardware design. The Ryzen AI Max+ 395 is a powerhouse compared to Intel’s Core Ultra 7 258V:
| Feature | AMD Ryzen AI Max+ 395 | Intel Core Ultra 7 258V |
||-|-|
| Cores / Threads | 16C / 32T | 8C / 8T |
| Max Boost Clock | 5.1 GHz | 4.8 GHz |
| AI Compute Power | 50+ AI TOPS (XDNA 2 NPU) | Lower AI optimization |
| TDP | 55W (configurable up to 120W) | 37W |
Clearly, AMD’s hardware is significantly more powerful, designed for high-end AI workloads, while Intel’s processor is optimized for power efficiency and lower-tier AI tasks.
Testing Conditions: Are They Fair?
Another important factor to consider is the test devices used in the benchmarks:
- AMD Ryzen AI Max+ 395 was tested on the Asus ROG Flow Z13 (64GB RAM, high-end gaming laptop).
- Intel Core Ultra 7 258V was tested on the Asus Zenbook S14 (32GB RAM, ultrabook with lower AI performance focus).
These differences in RAM capacity, power limits, and cooling systems make the comparisons less than ideal. The AMD-powered device had double the memory and a higher power budget, which likely contributed to its dominance in the benchmarks.
What Undercode Says: Analyzing the AI APU Market
AMD’s AI Push: Real Innovation or Just Marketing?
AMD’s strategy with the Ryzen AI Max+ 395 is clear: dominate AI workloads and position itself as the go-to choice for AI-focused devices. The XDNA 2 NPU and 50+ AI TOPS performance show that AMD is serious about AI acceleration, making this APU a compelling choice for LLM-based applications.
However, some skepticism remains. While AMD’s benchmarks demonstrate raw AI performance, real-world AI applications require a combination of hardware and software optimizations. Intel has historically had better AI software integration, especially through partnerships with Microsoft and enterprise solutions.
Is Intel Falling Behind?
Intel’s Core Ultra 7 258V is not designed to compete with the Ryzen AI Max+ 395 on AI performance. Instead, it focuses on energy efficiency and general computing. However, Intel has yet to release a true AI-focused APU that directly rivals AMD’s flagship.
If Intel doesn’t accelerate its AI hardware development, it risks losing significant ground in the AI laptop and mini PC markets.
Mini PCs and AI Computing: A New Frontier
Interestingly, AMD’s new APU is not just making waves in laptops. The mini PC market is rapidly adopting Ryzen AI hardware for compact AI workstations, edge computing, and gaming. Companies like GMKTec and Aoostar are leading the charge, pushing for more AI-driven mini PCs to hit the market between March and May 2025.
This shift suggests that AI-powered APUs are not just for high-end laptops—they are becoming central to compact computing as well.
Final Verdict: AMD’s Advantage is Clear, But Questions Remain
AMD’s Ryzen AI Max+ 395 is unquestionably more powerful than Intel’s Core Ultra 7 258V when it comes to AI workloads. However, the testing conditions and hardware disparities raise concerns about how fair the comparisons really are.
For users who prioritize AI performance—especially for LLMs, vision models, and data-heavy tasks—the Ryzen AI Max+ 395 is the best option right now. But for general users, the power efficiency and integration of Intel’s chips may still hold some advantages.
Fact Checker Results
- AMD’s Ryzen AI Max+ 395 is significantly more powerful in AI workloads, but direct comparisons with Intel are somewhat misleading due to hardware differences.
- Intel’s Core Ultra 7 258V is not a direct competitor; it prioritizes power efficiency and general computing rather than AI acceleration.
- AMD is leading the AI computing market, but Intel’s response (likely in future Meteor Lake or Lunar Lake chips) will determine if this lead is sustainable.
References:
Reported By: https://www.techradar.com/computing/cpu/the-amd-ryzen-ai-max-395-dominates-as-the-most-powerful-apu-on-the-market-but-its-competition-is-questionable
Extra Source Hub:
https://www.instagram.com
Wikipedia
Undercode AI
Image Source:
Pexels
Undercode AI DI v2
