gpurank
esc
  • Nothing named “”. Press Enter to search the offers instead.
↑↓ to move · ↵ to open · esc to close0 entries
Operator · Las Vegas, NV · founded 2023

TensorWave

US AMD-only AI cloud selling bare-metal AMD Instinct MI300X, MI325X and MI355X nodes (MI455X Helios racks announced) through sales-led contracts, with managed Kubernetes and Slurm.

GPU types3with a current offer
Current offers1111 readings on file · last 45 days
H100 from—per GPU-hour, on-demand
Entry price$1.71cheapest GPU-hour listed
Countries1regions not published
Trust Score v0.1
64/100
  • Company & track record100
  • Infrastructure control100
  • Pricing & billing transparency45
  • Reliability signals0
  • Security & compliance100
  • Customer signal—

Scored Sep 29, 2026 from the provider's own pages. Dimensions not yet measured are excluded from the denominator. Criteria ↓

Trust Score · 5 of 6 dimensions measured

The score, criterion by criterion

Each point comes from something the provider publishes, with the page it was read on and the date. Anything not disclosed earns zero.

Company & track record
weight 15%
100 / 100
Founding year published2023tensorwave.com · Sep 29, 20264 / 4
Headquarters publishedTensorWave, Inc.tensorwave.com · Sep 29, 20263 / 3
Legal entity namedTensorWave, Inc.tensorwave.com · Sep 29, 20264 / 4
Funding or ownership disclosedPrivate; states it has raised over USD 500M; site news: USD 350M Series B at USD 1.55B valuation (Jun 2026)tensorwave.com · Sep 29, 20264 / 4
Infrastructure control
weight 15%
100 / 100
Capacity model publishedoperates or rents dedicated capacitytensorwave.com · Sep 29, 202610 / 10
Regions published1 regions or countriestensorwave.com · Sep 29, 20265 / 5
Pricing & billing transparency
weight 20%
45 / 100
Public prices without an accountyes (read on the pricing page)tensorwave.com · Sep 29, 20265 / 5
Billing granularity publishedper hourtensorwave.com · Sep 29, 20264 / 4
Egress policy publishednot disclosedtensorwave.com · Sep 29, 20260 / 4
Stopped-instance billing publishednot disclosedtensorwave.com · Sep 29, 20260 / 4
Public IPv4 policy publishednot disclosedtensorwave.com · Sep 29, 20260 / 3
Reliability signals
weight 20%
0 / 100
Status pagenot disclosedtensorwave.com · Sep 29, 20260 / 6
SLA publishednot disclosedtensorwave.com · Sep 29, 20260 / 6
SLA with a numeric uptime commitmentnot disclosedtensorwave.com · Sep 29, 20260 / 3
Uptime history publishednot disclosedtensorwave.com · Sep 29, 20260 / 5
Observed availability (30 days)not measured yet—
Security & compliance
weight 15%
100 / 100
SOC 2Type IItensorwave.com · Sep 29, 20268 / 8
ISO 27001ISO 27001:2022 (2025 audit report on Trust Center)tensorwave.com · Sep 29, 20264 / 4
HIPAA or GDPR DPAcompliant (Trust Center)tensorwave.com · Sep 29, 20263 / 3
Customer signal
weight 15%
not measured
Aggregated public reviewsnot collected yet—
provisional v0.1: evidence from the provider's own pages; dimensions not yet measured are excluded from the denominatorFull method →
Prices by GPU · 3 GPU types with a current offer

What TensorWave charges, against the market

Cheapest current on-demand true cost here, next to the market median for the same GPU. Negative means cheaper than the market.

GPUCheapest hereMarket medianvs marketCurrent offersModes
MI300X$1.71$2.71−37%1on-demandMI300X prices
MI325X$2.25$2.65−15%1on-demandMI325X prices
MI355X$2.95$5.99−51%1on-demandMI355X prices
All offers · 11 readings on file

Every TensorWave offer, with its receipt

Current offers (last 45 days) ranked by true cost, all rental modes, each with its read date. Older readings stay listed below, in grey.

GPUConfigurationModeRegionStockListTrue costRead
MI300X8× MI300X2,300 GB RAM · 15,360 GB diskOn-demand—Stock unknown$1.71$1.7121:11 UTCfreshReceipt
MI325X8× MI325X3,000 GB RAM · 15,360 GB diskOn-demand—Stock unknown$2.25$2.2521:11 UTCfreshReceipt
MI355X8× MI355X3,072 GB RAM · 30,720 GB diskOn-demand—Stock unknown$2.95$2.9521:11 UTCfreshReceipt
3 fresh offers · prices per GPU-hour · true cost includes the add-ons the provider publishesOpen in the comparison tool →
True cost & billing · read Sep 29, 2026

What the hourly price does not say

Billing granularity, minimums, what a stopped machine costs, traffic, addresses, deposits and identity checks, as published by the provider.

Billing

Billing granularity
per hour
Minimum charge
not published
Stopped instance
not published
Egress
not published
Public IPv4
not published
Billing model
not published
Minimum deposit
not published
Free credits
not published

Account

Payment
crypto (not published)
Sign-up
sales contact form (company, company email, country, phone), sales contract
Identity check
sales contract
Support
email, chat, phone, support tickets, Discord
Enterprise support
not published
SLA
not published
Status page
not published
Published uptime
not published

Infrastructure

Regions
—
Countries
—
Platform
Jupyter, SSH, custom Docker images, persistent volumes, bare metalKubernetes: managedInterconnect: RoCEv2 3.2 Tb/s per node (MI300X, MI325X), 3 Tb/s (MI355X); AMD Infinity Fabric within node; AMD Pensando DPUsClusters: Bare-metal multi-node clusters with managed Slurm and managed Kubernetes; MI455X Helios rack-scale (up to 72 GPUs per rack) pricing TBD
Storage
not published
Migration out
Bare-metal Ubuntu 22.04 nodes with root access and Docker/Kubernetes/Slurm, so workloads are containers and can move to any ROCm host; moving to NVIDIA hardware requires ROCm-to-CUDA porting considerations. Contract terms govern exit.

Security & compliance

SOC 2
Type II
ISO 27001
ISO 27001:2022 (2025 audit report on Trust Center)
HIPAA
compliant (Trust Center)
GDPR / DPA
DPA page linked in site footer
Other
External, web app and internal pentest executive summaries 2026 on Trust Center
Legal entity
TensorWave, Inc.
Funding
Private; states it has raised over USD 500M; site news: USD 350M Series B at USD 1.55B valuation (Jun 2026)
Read on the provider's own pages · 15 pages read · facts marked "not published" were looked for and not foundPages read ↓
Assessment · documents read Sep 29, 2026

TensorWave, in short

What stood out when we read the documents: strengths, weaknesses, who it suits, and what we could not find.

Strengths

  • AMD Instinct specialist: MI300X from 1.71, MI325X from 2.25 and MI355X from 2.95 USD per GPU-hour published.
  • High-memory GPUs: 192 GB (MI300X), 256 GB (MI325X), 288 GB (MI355X) HBM per GPU.
  • Bare-metal 8-GPU nodes with root SSH access, 3.2 Tb/s RoCEv2 networking, managed Slurm and Kubernetes.
  • SOC 2 Type II, ISO 27001:2022 and HIPAA listed on the Trust Center.
  • Early access to next-generation AMD parts (MI455X Helios racks announced, pricing TBD).

Weaknesses

  • No self-serve sign-up: access via 'Book a call' form requiring company, company email and phone.
  • No NVIDIA GPUs.
  • Prices are 'Starting at' figures for 8-GPU nodes; term and commitment not published; docs FAQ says on-demand rental was only 'planned'.
  • No public pricing for storage, egress, IPs; no public SLA or status page found.
  • Data center locations not published.

Ideal for

  • AI teams committed to ROCm needing multi-node MI300X/MI325X/MI355X capacity under contract
  • Workloads bound by GPU memory capacity (large-model inference)

Not published

  • Mode and commitment term for the 'Starting at' prices not stated
  • Whether on-demand or hourly self-serve rental exists today not published (FAQ describing it as planned is about 2 years old).
  • No pricing page (tensorwave.com/pricing is 404)
  • no storage, egress or IPv4 prices.
  • Payment methods, billing granularity, deposits and identity verification not published.
  • Data center locations and countries not published.
  • API, CLI, SDK and Terraform availability not documented publicly.
  • SLA and status page not found.

Observations

tensorwave.com/pricing returns 'Page not found'; prices appear only on each accelerator product page as 'Starting at $X /GPU HR'

Every CTA leads to 'Book A Call' or 'Request POC'

MI455X listed with 'Pricing TBD'

Pages read

TensorWave · comparisons

Questions about TensorWave

What GPUs does TensorWave offer?

TensorWave has 3 GPU types with a fresh offer today, for example MI300X from $1.71, MI325X from $2.25, MI355X from $2.95 per GPU-hour on-demand. The full list with prices and market medians is on this page.

How does TensorWave bill?

Billing granularity: per hour.

How trustworthy is TensorWave?

Its Trust Score is 64/100 (v0.1), read from its own documents. Strong on company & track record, infrastructure control, security & compliance. Weaker on pricing & billing transparency, reliability signals. The score measures transparency and published guarantees, not customer satisfaction; every criterion and its source is listed above.

Where are TensorWave's data centres?

1 countries: —.

How do you pay TensorWave, and are there credits?

Payment by crypto. No sign-up credit is published. Identity check: sales contract.