The AI startup, Tiny Corp, has announced that it will be offering AI systems in both AMD & NVIDIA flavors with prices starting at $15K.
Tiny Corp Working With AMD To Resolve AI Software Optimizations, Now Offering NVIDIA Systems At $25K & AMD Systems Starting at $15K
We recently reported on the fiasco between AMD and Tiny Corp, where the AI firm demanded AMD to open-source some of its GPU firmware libraries to continue their development, but their request was rejected.
With that, TinyCorp previously announced its intentions to work with other GPU manufacturers, and now, finally, it has presented a new alternative, featuring six of NVIDIA's GeForce RTX 4090 GPUs powering the whole setup. However, the utilization of Team Green's products resulted in a massive price increase, as the NVIDIA Tiny Box solution reportedly costs $25,000, a whopping $10,000 increase from the AMD counterpart. With that said, the AMD system is also available with Radeon RX 7900 XTX GPUs and starts at $15K.
A hard to find "umr" repo has turned around the feasibility of the AMD tinybox. It will be a journey, but it gives us an ability to debug.
We're going to sell both, red for $15k and green for $25k. When you realize your preorder you'll choose your color. Website has been…
— the tiny corp (@__tinygrad__) March 25, 2024
Tiny Corp's decision to offer NVIDIA-powered AI systems does seem unfortunate for AMD's reputation among AI startups since refining the department of "software resources" is a critical one to progress in the markets, and NVIDIA hasn't left much of a gap for competitors to fill in, unlike AMD. Tiny Corp has expressed the fact that their company will still work with AMD to find a breakthrough, but their driver keeps on crashing in certain instances or completely stops working, according to the firm.
Tiny Corp simply states that if you want something that just works out of the box, then you should go for the NVIDIA GeForce-powered solution. They also use the famous quote from NVIDIA's CEO, Jensen Huang, "It Just Works" to represent the stability of NVIDIA's AI drivers.
Following are some of the features of Tiny Corp's Tinybox:
- 738 FP16 TFLOPs (AMD) / 991 FP16 TFLOPs (NVIDIA)
- 144 GB GPU RAM
- 5.76 TB/s (AMD) / 6.05 TB/s (NVIDIA) Ram Bandwidth
- 28.7 GB/s disk read bandwidth
- AMD EPYC CPU, 32 Cores
- 2x 1500W (two 120V Outlets, can power limit for less)
- Runs 70B FP16 LLaMA-2 out of the box using tinygrad
- $15,000 (AMD) / $25,000 (NVIDIA)
For those unaware, Tiny Corp's AI solution allows consumers to perform low-to-mid AI-focused workloads on dedicated machines without utilizing modern AI accelerators, which cost a hefty amount. As a solution, the company decided to package consumer GPUs into their hardware, due to their relatively more straightforward availability, lesser costs, and performance per dollar ratios as well. The company received immense interest from the markets and already received 583 preorders for the system, which were expected to start shipping by April.
tinybox with 6x Intel Arc A770 is colorful! pic.twitter.com/sM1GwIqnMn
— the tiny corp (@__tinygrad__) March 19, 2024
Interestingly, Tiny Corp has also created an Intel-based solution, mounting six of Intel's Arc A770, but the company doesn't plan to sell these anytime soon.
About the author: Muhammad Zuhair is a hardware and technology reporter for Wccftech, specializing in the semiconductor industry and the complex interplay between technology, manufacturing, and geopolitics. His coverage focuses on the corporate strategies and technological roadmaps of industry giants like TSMC, NVIDIA, Samsung, and Intel. Zuhair's expertise lies in deconstructing complex topics such as fabrication nodes (e.g., 2nm process), the economic impact of policies like the CHIPS Act, and the strategic development of AI infrastructure from NVIDIA, AMD and Intel.
Follow Wccftech on Google to get more of our news coverage in your feeds.
