DEPLOYDatabase

Chip comparison

Meta MTIA v2 vs NVIDIA L40S

Side-by-side specs

Straight from each vendor's datasheet. Winner in bold: higher throughput, memory, or transistor count; lower power draw or price. Sparse (2:4) throughput roughly doubles the FP8/INT8 numbers shown.

FieldMeta MTIA v2NVIDIA L40S
Process nodeTSMC N5TSMC 4N
Transistors (B)76.3
TDP90 W350 W
Memory128 GB LPDDR548 GB GDDR6
Memory bandwidth205 GB/s864 GB/s
FP16 (dense)362 TFLOPS
BF16 (dense)362 TFLOPS
FP8 (dense)733 TFLOPS
INT8 (dense)708 TOPS733 TOPS
Form factorPCIe (internal Meta)PCIe
Announced2024-04-102023-08-08
Released2024-05-01

Robots running Meta MTIA v2

No robots publicly running it yet.

Robots running NVIDIA L40S

No robots publicly running it yet.

Data centers with NVIDIA L40S

No data centers publicly running it yet.

Common questions

Meta MTIA v2 vs NVIDIA L40S: which has more memory?

Meta MTIA v2 carries more HBM (Meta MTIA v2: 128 GB LPDDR5; NVIDIA L40S: 48 GB GDDR6). Memory capacity sets the largest model that fits on one chip without partitioning; memory bandwidth sets inference throughput.

Meta MTIA v2 vs NVIDIA L40S: which is more widely used?

Meta MTIA v2: 0 robots and 2 data centers publicly running it. NVIDIA L40S: 0 robots and 0 data centers. Empty here means no operator has said publicly, not zero adoption in the market.

See also: every chip comparison · Meta MTIA v2 full page · NVIDIA L40S full page.