Jake Tech Lab

AMD MI355X vs Google TPU v7 Ironwood

AMD MI355X comes out ahead of Google TPU v7 Ironwood on 4 of the 6 things that decide this, going by what each maker publishes. The one that matters for you may be one the leader loses.

AMD MI355XWinner
AMD MI355X
7.3
Good
4 of 6 rounds
vs
Google TPU v7 Ironwood
Google TPU v7 Ironwood
7.4
Good
0 of 6 rounds

The verdict

AMD MI355X takes it, 4–0 on rounds.

On the published figures, AMD MI355X takes 4 of 6 rounds. Treat that as a starting point and check the current figures on both pages, because prices and plans move.

Choose AMD MI355X if…

  • • You want AMD MI355X's edge on how much memory it has and how fast its memory is

Choose Google TPU v7 Ironwood if…

  • • You have a reason to pick Google TPU v7 Ironwood that is not on this list

Round by round

  1. Round 1How much memory it hasAMD MI355X wins
    AMD MI355X8/10

    288 GB HBM3E

    dedicated memory size

    Source: amd.com ↗

    Google TPU v7 Ironwood5/10

    192 GB HBM

    hbm capacity per chip

    • 192 GB HBM, against 288 GB HBM3E

    Source: docs.cloud.google.com ↗

    AMD MI355X wins this round on what the company itself publishes.

  2. Round 2How fast its memory isAMD MI355X wins
    AMD MI355X8/10

    8 TB/s

    peak memory bandwidth

    Source: amd.com ↗

    Google TPU v7 Ironwood5/10

    7.37 TB/s

    hbm bandwidth per chip

    • 7.37 TB/s, against 8 TB/s

    Source: blog.google ↗

    AMD MI355X wins this round on what the company itself publishes.

  3. Round 3How fast it runs FP8 AI mathAMD MI355X wins
    AMD MI355X8/10

    5 PFLOPs

    peak ocp-fp8 matrix

    Source: amd.com ↗

    Google TPU v7 Ironwood5/10

    4,614 TFLOPs

    peak fp8 compute per chip

    • 4,614 TFLOPs, against 5 PFLOPs

    Source: blog.google ↗

    AMD MI355X wins this round on what the company itself publishes.

  4. Round 4How fast it runs FP16 AI mathAMD MI355X wins
    AMD MI355X8/10

    2.5 PFLOPs

    peak fp16 matrix

    Source: amd.com ↗

    Google TPU v7 Ironwood5/10

    2,307 TFLOPs

    BF16 peak per chip (Google publishes no FP16 figure)

    • 2,307 TFLOPs, against 2.5 PFLOPs

    Source: docs.cloud.google.com ↗

    AMD MI355X wins this round on what the company itself publishes.

  5. Round 5Which memory generation it usesTie
    AMD MI355X7/10

    HBM3E

    dedicated memory type

    Source: amd.com ↗

    Google TPU v7 Ironwood7/10

    HBM3E

    hbm type

    Source: cloud.google.com ↗

    Nothing to separate them here.

  6. Round 6How new it isTie
    AMD MI355X7/10

    6/12/2025

    launch date

    Source: amd.com ↗

    Google TPU v7 Ironwood7/10

    Apr 09, 2025

    announcement date

    Source: blog.google ↗

    Nothing to separate them here.

AMD MI355XAMD MI355X
Google TPU v7 IronwoodGoogle TPU v7 Ironwood

Side by side

AMD MI355XBest for how fast its memory isAMD MI355X
Overall score
7.3
7.4
Best
Price
See price
See price
How fast its memory is
7.9
7.3
How much memory it has
8.0
7.3
How much power it draws
5.0
—
How fast it runs FP16 AI math
8.0
7.3
How new it is
7.8
7.8
Head-to-head findings
How fast its memory is
8 TB/s
7.37 TB/s
How much memory it has
288 GB HBM3E
192 GB HBM
How much power it draws
1400W
—
How fast it runs FP16 AI math
2.5 PFLOPs
2,307 TFLOPs
How new it is
6/12/2025
Apr 09, 2025
Which memory generation it uses
HBM3E
HBM3E
How fast it runs FP8 AI math
5 PFLOPs
4,614 TFLOPs
How it handles scientific FP64 math
78.6 TFLOPs
—
Record
Head-to-head record
7–0
3–1
Head-to-heads won
100 %
75 %
Rounds won
69 %
75 %
Pros
    • peak bf16 compute per chip
    Cons
    • 1400W, against 700 W
    • 1400W, against 750W
    • 78.6 TFLOPs, against 163.4 TFLOPs
    • 1400W, against 1000W
    • 192 GB HBM, against 288 GB HBM3E
    • 7.37 TB/s, against 8 TB/s
    • 4,614 TFLOPs, against 5 PFLOPs
    • 2,307 TFLOPs, against 2.5 PFLOPs
    Add a third option to this comparison →

    More AI accelerators match-ups