DEPLOY

AI chip · Google

Google TPU v5p

Performance-tier TPU v5 generation, aimed at large-model training; ~2x compute vs TPU v4.

Deployed in 1 named data center · Part of 1 rack design.

Market position

v5p is the flagship of Google's fifth TPU generation, engineered for peak training throughput at 8,960-chip pod scale. Gemini 1.0 and 1.5 trained on it. v6p (Ironwood) is the announced successor; v6e (Trillium) is the inference-optimised sibling.

Efficiency and power

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

Perf per watt
1.53 BF16 TFLOPS/W
459 TFLOPS ÷ 300 W = 1.53
Cooling
water
Depends on the rack design

What fits in 95 GB

Which open-source LLMs run on one Google TPU v5p, 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).

ModelParamsFP16INT8INT4
Llama 3.1 8B
dense
8 B16 GB ✓8 GB ✓4 GB ✓
Llama 3.1 70B
dense
70 B140 GB ✗70 GB ✓35 GB ✓
Llama 3.1 405B
dense
405 B810 GB ✗405 GB ✗203 GB ✗
Llama 3.3 70B
dense
70 B140 GB ✗70 GB ✓35 GB ✓
DeepSeek V3
MoE (671B total, 37B active per token)
671 B1342 GB ✗671 GB ✗336 GB ✗
DeepSeek R1
MoE (671B total, 37B active per token)
671 B1342 GB ✗671 GB ✗336 GB ✗
Qwen 2.5 7B
dense
7 B14 GB ✓7 GB ✓4 GB ✓
Qwen 2.5 72B
dense
72 B144 GB ✗72 GB ✓36 GB ✓
Mixtral 8x7B
MoE (46.7B total, 12.9B active per token)
46.7 B93 GB ✓47 GB ✓23 GB ✓
Mixtral 8x22B
MoE (141B total, 39B active per token)
141 B282 GB ✗141 GB ✗71 GB ✓
Gemma 2 27B
dense
27 B54 GB ✓27 GB ✓14 GB ✓
Command R+
dense
104 B208 GB ✗104 GB ✗52 GB ✓
Kimi K2
MoE (1T total, 32B active per token)
1000 B2000 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 Google TPU v5p cost?

Google TPU v5p 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 Google TPU v5p have?

95 GB of HBM2e, running at 2,765 GB/s. Straight from the vendor datasheet. See chips with the most memory for context.

How much power does one Google TPU v5p draw?

300 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 data centers use Google TPU v5p?

1 named data center runs them, including Google The Dalles. See who has the most Google TPU v5p for the ranked list.

Which open-source LLMs fit on one Google TPU v5p?

Of 13 open-source LLMs we track, 4 fit at FP16, 7 at INT8, and 9 at INT4 (for example Llama 3.1 8B, Llama 3.1 70B and Llama 3.3 70B). Full table above with each model's memory need.

See every answer we publish →

Key facts

Class
Compute SoC (AI accelerator)
Designer
Google
Safety-critical?
No (data-center inference / training)
Record as of
2026-09-27
Most recent source
2024-01-01 (across 2 sources on this page)
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.

TDP
300 W
Memory
95 GB HBM2e
Memory bandwidth
2,765 GB/s
BF16 (dense)
459 TFLOPS
INT8 (dense)
918 TOPS
Form factor
OAM (per-chip in v5p pod)
Announced
2023-12-06
Released
2024-01-01

Source: vendor datasheet

Generation

Foundry & process

Fabbed at
TSMC (all chips TSMC makes →)

Rack designs that use it

Whole-rack systems buyers order in bulk (NVL72, HGX, TPU pods) that ship with Google TPU v5p inside.

See all rack designs →

Compare with

See every chip comparison →

Benchmarks

Published performance numbers by workload. Vendor datasheet figures where noted; independent measurements otherwise.

WorkloadValueUnitSourceAs of
peak tflops bf16459TFLOPSvendor2023-12-01

Data centers running Google TPU v5p

Who has the most? →

Named data-center campuses with Google TPU v5p on site. Counts shown where the operator has published them; other rows are described in general terms.

  1. partially energizedreported

    TPU v5p available in us-east5 (Columbus) + us-central1 (Iowa) + us-central2 (Oklahoma) — The Dalles hosts related pods

    Google Cloud

Sources

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 Google page.