$699 mini PC packs Nvidia Tesla V100 GPU for AI workloads
A $699 mini PC featuring an Nvidia Tesla V100 GPU has appeared on Newegg, targeting AI workloads with 16GB of HBM2 memory. The device’s pricing undercuts standalone V100 cards, but software limitations and seller risks may offset its appeal for founders and operators.
Editor, Lazyfounder
TABLE OF CONTENTS
What the Rmatamini P99S offers
Newegg is listing a compact AI mini PC, the Rmatamini P99S, for $699. The device includes an Nvidia Tesla V100 GPU with 16GB of HBM2 memory, an Intel Core i7-10850H processor, 16GB of RAM, and a 512GB SSD. The system ships from China and runs Windows 11 Pro.
Why the Tesla V100’s specs still matter
The Tesla V100 GPU, once sold for over $10,000, is now available in the P99S for a fraction of its original price. Its 900GB/s memory bandwidth remains competitive, nearly double that of an RTX 5060 Ti 16GB. This makes it appealing for AI workloads that rely on high memory throughput, such as token generation in large language models.
Standalone 16GB Tesla V100 GPUs sell for $584 new on Amazon, with renewed units available for $410. The P99S’s $699 price tag suggests the rest of the system—including the CPU, RAM, and SSD—costs between $115 and $290, depending on the GPU’s valuation.
AI memory costs have risen sharply this year, driven by demand for high-bandwidth solutions. Nvidia’s DGX Spark price increase in February, coupled with global shortages, has made alternatives like the V100 more attractive despite its age.
Software and memory limitations
The P99S’s pricing is aggressive, but buyers face critical trade-offs. Nvidia’s R580 driver branch is the last to support the V100, and CUDA 13 no longer targets it. Community tools like Ollama and llama.cpp still work on CUDA 12, but newer frameworks, such as vLLM, require compute capabilities beyond what the V100 offers.
The GPU’s tensor cores support only FP16, limiting its utility for startups relying on newer low-precision formats like BF16. This could pose challenges for training or fine-tuning models optimized for modern hardware.
Memory constraints are another hurdle. The 16GB HBM2 can handle mixture-of-experts models like Gemma 4 26B-A4B, but larger models may require overflow to system RAM, which is also limited to 16GB in the base configuration. Rmatamini offers a $1,500 SKU with 32GB of RAM for heavier workloads.
Hardware and seller risks
The Core i7-10850H, a six-core mobile processor from 2020, matches the V100’s PCIe 3.0 interface but offers no display outputs—the GPU itself lacks video ports, so the system relies on the CPU’s integrated graphics.
The seller, Rmatamini, has only one rating, and the product itself has none. Warranty support is unclear, with returns limited to a 30-day refund window. Founders evaluating this for prototyping or edge deployments should account for potential compatibility or support issues.
Related on Lazyfounder
Sources
- TechRadar · 2026-11-09
You can grab an AI mini PC housing an Nvidia Tesla GPU with 16GB of HBM2 for just $699
This story is an original summary drafted with AI by Lazyfounder from the reporting listed above and checked by automated validation. Facts are attributed to their original publishers; sections marked as analysis are Lazyfounder's. Where a source is in another language, facts were machine-translated and quotations are reported, not reproduced. Read the original coverage via the links, and see our AI policy and corrections policy.
About the author
Editor, Lazyfounder
Tarun Mottlia edits LazyFounders, covering Indian startups, funding rounds, AI and product launches. Every story on the site is AI-assisted and checked against its cited sources before publication.
More stories by Tarun MottliaGet the LazyFounder Brief
Startup, funding and AI news in a five-minute read. Join the early-access list.
