Tech giants Ampere and Qualcomm are joining forces to create a new generation of AI servers designed for efficiency and scalability. This collaboration combines Ampere’s high-performance, low-power server CPUs with Qualcomm’s Cloud AI 100 Ultra AI inferencing chips.
The focus of these new servers is on running existing AI models, rather than training them from scratch. This “inference” stage is crucial for real-world applications of AI, such as facial recognition or language translation.
Ampere’s efficiency edge and scaling up AI with Qualcomm
Ampere has carved a niche for itself by offering server chips that deliver strong performance while consuming minimal power. This focus on efficiency makes them ideal for data centers, where energy costs can be significant. While Ampere’s CPUs excel at general-purpose tasks and basic AI inference, Qualcomm‘s AI chips are designed to handle much larger and more complex models.
This partnership allows Ampere to offer a complete solution for data centers running ever-more demanding AI applications. As part of this collaboration, Ampere also announced its latest generation of server chips, the AmpereOne. These chips boast a massive 256 cores and cutting-edge 3nm manufacturing process. Additionally, they feature 12-channel DDR5 memory, allowing for even more efficient data access.
Ampere emphasizes that raw performance is just one piece of the puzzle. By focusing on lower power consumption, they aim to reduce the total cost of ownership for data centers running AI workloads. This is particularly important when compared to competing solutions from Nvidia. This Qualcomm partnership isn’t Ampere’s only collaboration.
The company also announced a joint project with NETINT to create servers specifically designed for video processing. These servers can handle demanding tasks like transcoding 360-degree live video streams while simultaneously generating subtitles using AI-powered speech recognition.
Ampere’s CEO, Renee James, reiterated the company’s core mission: to deliver high-performance computing solutions without sacrificing on efficiency. They believe they’ve proven that low power consumption doesn’t have to equate to low performance. This new wave of AI servers from Ampere and Qualcomm seems promising enough in offering businesses a powerful and cost-effective way to leverage the power of artificial intelligence.
