DEPLOYDatabase

Chip comparison

AMD Instinct MI300X vs Google TPU v5e

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 MI300XGoogle TPU v5e
Process nodeTSMC N5 + N6 (chiplet)
Transistors (B)153
TDP750 W170 W
Memory192 GB HBM316 GB HBM2
Memory bandwidth5,300 GB/s819 GB/s
FP16 (dense)1,307 TFLOPS
BF16 (dense)1,307 TFLOPS197 TFLOPS
FP8 (dense)2,614 TFLOPS
INT8 (dense)2,614 TOPS393 TOPS
Launch price (list)$15,000
Form factorOAMOAM (v5e pod, up to 256 chips)
Announced2023-12-062023-08-29
Released2024-01-012023-11-01

Robots running AMD Instinct MI300X

No robots publicly running it yet.

Robots running Google TPU v5e

No robots publicly running it yet.

Data centers with AMD Instinct MI300X

No data centers publicly running it yet.

Common questions

AMD Instinct MI300X vs Google TPU v5e: which is faster for training?

AMD Instinct MI300X has 13.27x the dense FP16/BF16 throughput of the other (AMD Instinct MI300X: 2,614 TFLOPS; Google TPU v5e: 197 TFLOPS). Real-world training speed also depends on memory bandwidth, interconnect, and software stack: see the spec diff for those inputs.

AMD Instinct MI300X vs Google TPU v5e: which has more memory?

AMD Instinct MI300X carries more HBM (AMD Instinct MI300X: 192 GB HBM3; Google TPU v5e: 16 GB HBM2). Memory capacity sets the largest model that fits on one chip without partitioning; memory bandwidth sets inference throughput.

AMD Instinct MI300X vs Google TPU v5e: which is more power-efficient?

Dense FP16/BF16 TFLOPS per watt: AMD Instinct MI300X 3.49 (2,614 ÷ 750 W); Google TPU v5e 1.16 (197 ÷ 170 W). AMD Instinct MI300X wins at the die level; system-level efficiency also depends on cooling and interconnect.

AMD Instinct MI300X vs Google TPU v5e: which is more widely used?

AMD Instinct MI300X: 0 robots and 0 data centers publicly running it. Google TPU v5e: 0 robots and 1 data center. Empty here means no operator has said publicly, not zero adoption in the market.

See also: every chip comparison · AMD Instinct MI300X full page · Google TPU v5e full page.