DEPLOYDatabase

Chip comparison

AMD Instinct MI350 series vs Google TPU v4

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.

FieldAMD Instinct MI350 seriesGoogle TPU v4
Process nodeTSMC N3 (chiplet)TSMC N7
TDP1000 W192 W
Memory288 GB HBM3e32 GB HBM2
Memory bandwidth8,000 GB/s1,200 GB/s
FP16 (dense)2,300 TFLOPS
BF16 (dense)2,300 TFLOPS275 TFLOPS
FP8 (dense)4,600 TFLOPS
INT8 (dense)4,600 TOPS275 TOPS
Form factorOAMOAM (per-chip in v4 pod)
Announced2025-06-122021-05-18
Released2025-10-012022-05-01

Robots running AMD Instinct MI350 series

No robots publicly running it yet.

Robots running Google TPU v4

No robots publicly running it yet.

Data centers with AMD Instinct MI350 series

Common questions

AMD Instinct MI350 series vs Google TPU v4: which is faster for training?

AMD Instinct MI350 series has 16.73x the dense FP16/BF16 throughput of the other (AMD Instinct MI350 series: 4,600 TFLOPS; Google TPU v4: 275 TFLOPS). Real-world training speed also depends on memory bandwidth, interconnect, and software stack: see the spec diff for those inputs.

AMD Instinct MI350 series vs Google TPU v4: which has more memory?

AMD Instinct MI350 series carries more HBM (AMD Instinct MI350 series: 288 GB HBM3e; Google TPU v4: 32 GB HBM2). Memory capacity sets the largest model that fits on one chip without partitioning; memory bandwidth sets inference throughput.

AMD Instinct MI350 series vs Google TPU v4: which is more power-efficient?

Dense FP16/BF16 TFLOPS per watt: AMD Instinct MI350 series 4.60 (4,600 ÷ 1000 W); Google TPU v4 1.43 (275 ÷ 192 W). AMD Instinct MI350 series wins at the die level; system-level efficiency also depends on cooling and interconnect.

AMD Instinct MI350 series vs Google TPU v4: which is more widely used?

AMD Instinct MI350 series: 0 robots and 1 data center publicly running it. Google TPU v4: 0 robots and 2 data centers. Empty here means no operator has said publicly, not zero adoption in the market.

See also: every chip comparison · AMD Instinct MI350 series full page · Google TPU v4 full page.