DEPLOYDatabase

Chip comparison

Google TPU v4 vs Intel Gaudi 2

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.

FieldGoogle TPU v4Intel Gaudi 2
Process nodeTSMC N7TSMC N7
TDP192 W600 W
Memory32 GB HBM296 GB HBM2e
Memory bandwidth1,200 GB/s2,450 GB/s
BF16 (dense)275 TFLOPS432 TFLOPS
INT8 (dense)275 TOPS—
Form factorOAM (per-chip in v4 pod)OAM
Announced2021-05-182022-05-10
Released2022-05-012022-09-01

Robots running Google TPU v4

No robots publicly running it yet.

Robots running Intel Gaudi 2

No robots publicly running it yet.

Data centers with Intel Gaudi 2

No data centers publicly running it yet.

Common questions

Google TPU v4 vs Intel Gaudi 2: which is faster for training?

Intel Gaudi 2 has 1.57x the dense FP16/BF16 throughput of the other (Google TPU v4: 275 TFLOPS; Intel Gaudi 2: 432 TFLOPS). Real-world training speed also depends on memory bandwidth, interconnect, and software stack: see the spec diff for those inputs.

Google TPU v4 vs Intel Gaudi 2: which has more memory?

Intel Gaudi 2 carries more HBM (Google TPU v4: 32 GB HBM2; Intel Gaudi 2: 96 GB HBM2e). Memory capacity sets the largest model that fits on one chip without partitioning; memory bandwidth sets inference throughput.

Google TPU v4 vs Intel Gaudi 2: which is more power-efficient?

Dense FP16/BF16 TFLOPS per watt: Google TPU v4 1.43 (275 ÷ 192 W); Intel Gaudi 2 0.72 (432 ÷ 600 W). Google TPU v4 wins at the die level; system-level efficiency also depends on cooling and interconnect.

Google TPU v4 vs Intel Gaudi 2: which is more widely used?

Google TPU v4: 0 robots and 2 data centers publicly running it. Intel Gaudi 2: 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 · Google TPU v4 full page · Intel Gaudi 2 full page.