AI chip · AMD
AMD Instinct MI350 series
CDNA 4 next-generation accelerator (MI350X, MI355X) with 288 GB HBM3e and FP4/FP6 support. Announced 2024; ramping through 2025-2026.
Deployed in 1 named data center.
Market position
AMD's answer to Blackwell B200. 288 GB HBM3e, positioned against B200's 192 GB. HUMAIN, Meta, Oracle and Microsoft are the announced buyers. Successor MI400 lands 2026 on a rack-scale design (Helios) against GB200 NVL72.
Efficiency and power
How much work you get per watt and per dollar, and how much power a full rack draws.
- Perf per watt
- 4.60 FP8 TFLOPS/W4,600 TFLOPS ÷ 1000 W = 4.60
What fits in 288 GB
Which open-source LLMs run on one AMD Instinct MI350 series, by precision. Weights only: add roughly 15% headroom for real serving. If a model does not fit at FP16, try INT8 or INT4 (smaller quality trade-off than most people expect).
| Model | Params | FP16 | INT8 | INT4 |
|---|---|---|---|---|
| Llama 3.1 8B dense | 8 B | 16 GB ✓ | 8 GB ✓ | 4 GB ✓ |
| Llama 3.1 70B dense | 70 B | 140 GB ✓ | 70 GB ✓ | 35 GB ✓ |
| Llama 3.1 405B dense | 405 B | 810 GB ✗ | 405 GB ✗ | 203 GB ✓ |
| Llama 3.3 70B dense | 70 B | 140 GB ✓ | 70 GB ✓ | 35 GB ✓ |
| DeepSeek V3 MoE (671B total, 37B active per token) | 671 B | 1342 GB ✗ | 671 GB ✗ | 336 GB ✗ |
| DeepSeek R1 MoE (671B total, 37B active per token) | 671 B | 1342 GB ✗ | 671 GB ✗ | 336 GB ✗ |
| Qwen 2.5 7B dense | 7 B | 14 GB ✓ | 7 GB ✓ | 4 GB ✓ |
| Qwen 2.5 72B dense | 72 B | 144 GB ✓ | 72 GB ✓ | 36 GB ✓ |
| Mixtral 8x7B MoE (46.7B total, 12.9B active per token) | 46.7 B | 93 GB ✓ | 47 GB ✓ | 23 GB ✓ |
| Mixtral 8x22B MoE (141B total, 39B active per token) | 141 B | 282 GB ✓ | 141 GB ✓ | 71 GB ✓ |
| Gemma 2 27B dense | 27 B | 54 GB ✓ | 27 GB ✓ | 14 GB ✓ |
| Command R+ dense | 104 B | 208 GB ✓ | 104 GB ✓ | 52 GB ✓ |
| Kimi K2 MoE (1T total, 32B active per token) | 1000 B | 2000 GB ✗ | 1000 GB ✗ | 500 GB ✗ |
Math: FP16 = params × 2 bytes; INT8 = params × 1 byte; INT4 = params × 0.5 bytes. MoE models sum every expert (full weights on disk), not the per-token active subset.
Common questions
How much does AMD Instinct MI350 series cost?
AMD Instinct MI350 series doesn't have a public launch list price. Vendors like this one usually price through direct sales rather than a public sheet. See list price per TFLOP for chips that do publish.
How much memory does AMD Instinct MI350 series have?
288 GB of HBM3e, running at 8,000 GB/s. Straight from the vendor datasheet. See chips with the most memory for context.
How much power does one AMD Instinct MI350 series draw?
1000 W at the chip. A full server draws more once you add CPU, memory, networking and cooling: see the rack power number above. Compare to other chips on perf-per-watt.
Which data centers use AMD Instinct MI350 series?
1 named data center runs them, including HUMAIN (Riyadh & Dammam flagship data centers). See who has the most AMD Instinct MI350 series for the ranked list.
Which open-source LLMs fit on one AMD Instinct MI350 series?
Of 13 open-source LLMs we track, 9 fit at FP16, 9 at INT8, and 10 at INT4 (for example Llama 3.1 8B, Llama 3.1 70B and Llama 3.1 405B). Full table above with each model's memory need.
Key facts
- Class
- Compute SoC (AI accelerator)
- Designer
- AMD
- Safety-critical?
- No (data-center inference / training)
- Record as of
- 2026-09-27
- Most recent source
- 2025-05-13 (across 1 source on this page)
Specifications (11 fields, click to expand)
Straight from the vendor datasheet. Dense throughput shown first; sparse (2:4) numbers in parentheses where the vendor publishes them. Full datasheet linked below.
- Process node
- TSMC N3 (chiplet)
- TDP
- 1000 W
- Memory
- 288 GB HBM3e
- Memory bandwidth
- 8,000 GB/s
- FP16 (dense)
- 2,300 TFLOPS
- FP8 (dense)
- 4,600 TFLOPS
- FP4 (dense)
- 10,100 TFLOPS (20,100 sparse)
- INT8 (dense)
- 4,600 TOPS
- Form factor
- OAM
- Announced
- 2025-06-12
- Released
- 2025-10-01
Source: vendor datasheet
Generation
Foundry & process
- Fabbed at
- TSMC (all chips TSMC makes →)
- Process node
- TSMC N3 (chiplet)
Compare with
Data centers running AMD Instinct MI350 series
Who has the most? →Named data-center campuses with AMD Instinct MI350 series on site. Counts shown where the operator has published them; other rows are described in general terms.
- HUMAIN (Riyadh & Dammam flagship data centers)as of 2025-05-13under constructionreported
AMD Instinct MI350 series committed for HUMAIN AI factories
Sources
Adoption rows appear as we confirm each chip-in-robot pairing from a public source. See every chip we track for the full catalog or the AMD page.