Advancing AI: Polywell's Latest Innovations in Cloud and Edge Computing
Polywell Computers continues to be at the forefront of technological innovation, particularly in the realm of AI computing. As we move forward, our commitment to supporting both Cloud and Edge AI remains unwavering, building on the foundations we discussed in our previous blog post from November 2023.
In that post, we highlighted the growing competition between Nvidia’s Cuda platform and Intel’s OpenVino, both of which are driving significant advancements in AI. We also touched on how Nvidia’s A100 GPU has become a leader in high-end AI computing, and how Intel’s Gaudi processors are shaping up to be formidable competitors. These trends continue to shape our strategy as we expand our product offerings to meet the evolving needs of AI developers.
Today, we’re excited to share more about how Polywell is adapting to the challenges and opportunities in this rapidly evolving field. As many of you know, the Nvidia A100 remains a premium choice for AI workloads, but it’s becoming increasingly difficult and expensive to source. Recognizing this, we’ve broadened our product line to include solutions that leverage the more readily available RTX4090 GPU. This allows developers to continue pushing the boundaries of AI without being constrained by hardware shortages.
Our new 5U servers, designed to accommodate up to eight RTX4090 GPUs, are a testament to our commitment to providing flexible and powerful computing solutions. These systems are perfect for those who need substantial computational power without the high costs associated with premium GPUs like the A100.
5U Server with 8 RTX4090 GPU
5U Dual AMD 7002/7003 EPYC™,
8 GPU + 2 PCIe 4.0 x16, 8 2.5” + M.2 NVMe,
32 DIMM 8-Channel DDR4, 2 10GbE LAN + IPMI
5U Dual Socket 3rd Gen. Intel® Xeon® CPU,
8 GPU + 2 PCIe 4.0 x16, 8 2.5” + M.2 NVMe,
32 DIMM 8-Channel DDR4, 2 10GbE LAN + IPMI
For mainstream AI computing, our 4U/Tower Xeon configurations, supporting dual RTX4090 GPUs, offer a balanced approach between performance and cost-effectiveness.
4U / Tower Xeon with 2 RTX4090 for Mainstream AI Computing
GPU Computing for AI
4th Gen. Xeon W-3400
8-ch. DDR 5.0
4 x PCIe 5.0 x16 Slots
Up to 4 Quad Nvidia Graphics
or Dual RTX-4090 GPU Cards
Building on the discussion from our previous blog, Nvidia continues to dominate the AI GPU market with its A100, and Polywell remains fully committed to supporting this platform. Our 2U and 4U systems are optimized to take full advantage of the A100’s capabilities, ensuring that our customers have access to the best hardware for their AI needs.
However, we’re also keeping a close eye on Intel’s developments. As we mentioned in November, Intel is poised to challenge Nvidia’s dominance with its Gaudi processors and ARC graphics cards. We are confident that Intel’s advancements in PCIe 5.0 and 6.0, coupled with their robust software ecosystem, will position them as strong competitors in the AI space. At Polywell, we’re prepared to integrate these technologies into our future product lines, betting on Intel’s ability to reclaim its place as a leader in AI computing.
Intel OpenVino SFF platform:
14th Gen. Meteor Lake,Intel® Arc™ GPU,
Dual Display HDMI + DP,
M.2 NVMe SSD with 4 USB + Type-C,
Digital Signage Player PC
Our focus on AI isn’t limited to the cloud; we’re also expanding our offerings in Edge AI. The AI Edge PCs we discussed previously are now being updated with the latest Nvidia Cuda and Intel OpenVino platforms, ensuring that our customers have access to the most advanced hardware for edge computing. These compact systems are ideal for applications that require immediate processing power at the data source, offering a perfect blend of performance and efficiency. For AI Edge PC, we have i2317-RTX4060 as new generation extended from the i2314-RTX3050.
Nvidia powered SFF Edge AI PC:
13th Gen. Intel® H-Series i9/i7,
NVIDIA® RTX4060 Graphics, 2 HDMI + Type-C,
3 M.2 slots + 4 USB 3.0 + 4 USB 2.0
Beyond AI, Polywell continues to innovate in the embedded computing space with our Mini PCs and ARM Android platforms. We’re also expanding our server line to support the latest advancements in GPU and network I/O technologies, including 200G and 400G Network I/O modules through PCIe 5.0 and PCIe 6.0.
We are going to add the new generation of Ultra Core family in Q4 2024 which includes the H810/B860/Q870/W880 based Socket-S family, u/H/V-series of Mobile processor family like Meteor Lake, Lunar Lake and Arrow Lake.
For those interested in the detailed specifications of our new products, we encourage you to visit our website, where you can find comprehensive information about each system.
As always, Polywell is committed to providing cutting-edge technology that empowers our customers to achieve their goals, whether in AI, embedded computing, or beyond.
We look forward to continuing this journey with you, building on the solid foundation we’ve established and embracing the exciting challenges that lie ahead.
Stay tuned for more updates, and as always, feel free to reach out with any questions.
Best regards,
Sam Chu