AMD Instinct MI300X

What hackers pick instead

• NVIDIA H100People recommend it for top AI training and cheap rental access, though critics cite high purchase cost, power draw and failure rates.20recommend
• NVIDIA Tesla V100People recommend the used V100 as cheap VRAM and FP64 compute for local AI and physics work. Caveats: no bfloat16, loud fans, high idle power.16recommend
• AMD Instinct MI355XStrong memory and performance per dollar for large AI inference and training; critics say its software stack lags and it trails Nvidia's B300 on some benchmarks.14recommend
• Nvidia A100Long-lived datacenter GPU that people still run for AI and compute work years after release; the most common caveat is that it can be unreliable under long runs.10recommend
• Intel D270 GPU AI AcceleratorRecommended for holding many resident AI models in 480 GB of LPDDR5X memory. Critics say its low memory bandwidth makes it slow, and it is very expensive for individuals.10recommend

Every quote

The 15 most detailed of 15 comments. Each links to the original.

1. it's a datacenter gpu, and one that actually has more VRAM than the NVIDIA equivalent

   Recommends · Aug 2026 · in “DeepSeek V4 Flash on a Single AMD MI300X”

2. MI300X is quite accessible to a small team with budget for just 1-2 GPUs

   Recommends · Aug 2026 · in “DeepSeek V4 Flash on a Single AMD MI300X”

3. The MI300X will vastly outperform it for only a slightly higher price.

   Recommends · Aug 2026 · in “DeepSeek V4 Flash on a Single AMD MI300X”

4. The Mi300 is very competitive on token throughput as far as I'm aware.

   Recommends · Nov 2025 · in “Michael Burry a.k.a. "Big Short",discloses $1.1B bet against Nvidia&Palantir”

5. the higher HBM on Mi300x is really useful for this kind of thing

   Recommends · Aug 2026 · in “DeepSeek V4 Flash on a Single AMD MI300X”

6. multiple boxes of 8x AMD MI300x available right now

   Recommends · Jan 2026 · in “AWS raises GPU prices 15% on a Saturday, hopes you weren't paying attention”

7. The only one even close, is AMD.

   Recommends · Jun 2026 · in “Bringing Up DeepSeek-V4-Flash on AMD MI300X”

8. the fact you cannot even do some very basic stuff on AMD is going to mean you are left unused by researchers

   Criticizes · Apr 2026 · in “Why isn't AMD's MI300X competitive?”

9. int4 quantization not supported. fp8 quantization not supported.

   Criticizes · Apr 2026 · in “Microsoft to Stop Sharing Revenue with Main AI Partner OpenAI”

10. nowhere near the same software stack, so nobody cares

    Criticizes · Mar 2026 · in “Cloud VM benchmarks 2026: performance/price for 44 VM types over 7 providers”

11. I ran nanochat on MI300X and it just worked.

    Recommends · Nov 2025 · in “HipKittens: Fast and furious AMD kernels”

12. hardware is okay (potentially great?!)

    Mixed · Apr 2026 · in “Why isn't AMD's MI300X competitive?”

13. I still have nightmares of the L3 cache cross-IOD bottlenecks on MI300X

    Criticizes · May 2026 · in “Real-time LLM Inference on Standard GPUs: 3k tokens/s per request”

14. the MI300x didn't replace H100s en masse.

    Criticizes · Dec 2025 · in “Asahi Linux with Sway on the MacBook Air M2”

15. This is not even remotely worth it.

    Criticizes · Aug 2026 · in “DeepSeek V4 Flash on a Single AMD MI300X”