SambaNova SN40L

Reconfigurable Dataflow Unit optimized for enterprise LLM serving with large-parameter memory hierarchies.

Market position

SambaNova's Reconfigurable Dataflow Unit (RDU). 1.5 TB tier-3 memory per socket enables trillion-parameter models on 8 chips. Serves Samba-1 Composition of Experts. Used by SoftBank, Analog Devices, ArgonneNational Lab. Positioned against DGX H100 SuperPOD on memory-bound inference.

Efficiency and power

How much work you get per watt and per dollar, and how much power a full rack draws.

Perf per watt: 1.79 FP8 TFLOPS/W
1,250 TFLOPS ÷ 700 W = 1.79

What fits in 1500 GB

Which open-source LLMs run on one SambaNova SN40L, 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 SambaNova SN40L cost?

SambaNova SN40L 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 SambaNova SN40L have?

1500 GB of DDR5 (three-tier: 64 MB SRAM + 64 GB HBM3 + 1.5 TB DDR5). Straight from the vendor datasheet. See chips with the most memory for context.

How much power does one SambaNova SN40L draw?

700 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 open-source LLMs fit on one SambaNova SN40L?

Of 13 open-source LLMs we track, 12 fit at FP16, 13 at INT8, and 13 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.

See every answer we publish →

Key facts

Class: Compute SoC (AI accelerator)
Designer: SambaNova Systems
Safety-critical?: No (data-center inference / training)
Record as of: 2026-09-27

Specifications (8 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 N5
TDP: 700 W
Memory: 1500 GB DDR5 (three-tier: 64 MB SRAM + 64 GB HBM3 + 1.5 TB DDR5)
FP16 (dense): 638 TFLOPS
FP8 (dense): 1,250 TFLOPS
Form factor: PCIe (SN40L DataScale system)
Announced: 2023-09-19
Released: 2024-01-01

Source: vendor datasheet

Foundry & process

Fabbed at: TSMC (all chips TSMC makes →)
Process node: TSMC N5

Sources

SambaNova SN40L RDUSambaNova

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 SambaNova Systems page.