AMD’s MI300X Is a Far Superior Option Than NVIDIA’s H100s, Says AI Startup TensorWave’s CEO

Apr 17, 2024 at 03:15am EDT
AMD MI300X Up To 3x Faster Than NVIDIA H100 In LLM Inference AI Benchmarks, Offers Competitive Pricing Too 1

AI startup TensorWave, believes that AMD's Instinct MI300X accelerator is a far better option than competitors such as NVIDIA's H100.

The Availability of AMD's Instinct MI300X Accelerator, Coupled With The Better Price-To-Performance Ratio, Puts It Ahead of NVIDIA, According To AI Startup's CEO

Note - Do check out the full blog and guide by Tensorwave on how AMD's Instinct MI300X is being used for LLM inferencing and outperforms NVIDIA's H100 platform.

Related Story NVIDIA Floods Europe With 35 Supercomputers Spanning 23 Countries, Stacking Up To 800 Exaflops Of AI Compute

AMD has gained massive popularity in just six months, with the firm receiving massive orders from the industry for its flagship Instinct accelerator. While the Instinct MI300X has existed for a while in the markets, the factor that has made it a more compelling option is how AMD has ensured a more steady availability of the accelerator, not to mention the "waves" of upgrades we have witnessed within AMD's software ecosystem, which has taken the performance to a whole new level.

TensorWave's Co-Founder Jeff Tatarchuk sees AMD's Instinct MI300X accelerator as a far superior option to NVIDIA's H100s. He says they are "available to buy," unlike NVIDIA, which still witnesses order backlogs with AI GPUs.

TensorWave is working on providing its extensive arsenal of AMD's MI300X AI accelerator of 20,000 units for leasing at a fraction of the cost of what the original SKU is priced. Moreover, the firm plans to rapidly accelerate AMD's adoption in the markets through widespread advertisements & ensure its consumers that AMD is the more viable choice in modern times.

Apart from TensorWave's interest, we recently reported that the MI300X is causing headaches for competitors out there. While the accelerator hasn't received the level of attention NVIDIA attained with its Hopper lineup, we certainly feel the future is bright for AMD.

NVIDIA did a great job in establishing a market image. Still, in the performance department, based on widespread research, MI300X dominates in various aspects, whether it is memory capacities or compute power, which shows that AMD could be in a competitive spot moving ahead, given that they manage to grow the ecosystem as a whole.

With the debut of NVIDIA's Blackwell, the industry is siding with the firm right now. Still, if AMD's next-generation accelerators, the MI400s, manage to overtake Blackwell in every department, AMD could very well be in an AI market bull run in the upcoming years.

News Source: The Register

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.