TensorDock
Search…
⌃K

Data Centers

This is where your data gets stored

Compute

These are where compute workloads on TensorDock run.

Boston DC1, Our Mining Farm (bos1)

  • Reservable instances
  • No data center tier
  • NVIDIA GeForce GPUs
Our Boston 01 location primarily houses GeForce cards for Ethereum mining. It is single-homed onto Verizon and has no power or cooling redundancy. All hardware here, from the network switches to the racks to the GPUs, is fully owned by us. This location is located in the Greater Boston Area.
For customer looking for dirt-cheap compute on monthly or longer commitments, we can provide heavily-discounted compute servers from here.
Deployed GPUs:
  • GeForce GTX/RTX (1070, 1650, 2080S, 3060Ti, 3070, 3070)
iPerf Speedtest: bos1.speedtest.tensordock.com

Boston DC2, Our Headquarters (bos2)

  • Reservable instances
  • Tier 1 design
  • NVIDIA GeForce GPUs, NVIDIA Enterprise GPUs
Our Boston 02 location houses a number of NVIDIA Enterprise cards and NVIDIA GeForce cards as well. We are single homed onto Verizon Business. All hardware here, from the network switches to the racks to the GPUs, is fully owned by us. We have our own lease for the 2 floors that we rent (one for servers, another for part storage/office). We have cooling redundancy and UPSes for power outages. In addition, our servers run on A+B power, though the actual circuits come from the same power upstream. This is primarily used for our hardware R&D and hosts a number of cost-sensitive customers as well. This location is located in the Greater Boston Area.
For customer looking for affordable compute on monthly or longer commitments, we can provide discounted compute servers from here.
Deployed GPUs:
  • GeForce GTX/RTX (1070, 1080, 1650, 1660S, 2070S, 2080, 2080S, 2080Ti, 3080)
  • NVIDIA A-series
iPerf Speedtest: bos2.speedtest.tensordock.com

New York, New Jersey DC1 (ewr1)

  • On-demand instances, reservable instances
  • Tier 2 design
  • NVIDIA Enterprise GPUs
This New York location, which is actually located in New Jersey, provides a limited number of on-demand GPUs.
Deployed GPUs:
  • V100 SXM
  • V100 PCIE
  • Quadro RTX 4000
  • Quadro RTX 5000
iPerf Speedtest: ewr1.speedtest.tensordock.com

New York, Long Island DC1 (lga1)

  • Reservable instances
  • Tier 2 design
  • NVIDIA Enterprise GPUs
This New York location, which is actually located on Long Island, provides a limited number of on-demand GPUs.
Deployed GPUs:
  • RTX A4000
  • Quadro RTX 4000
  • Quadro RTX 5000

Chicago DC1, SBA Edge (ord1)

  • On-demand instances, reservable instances
  • Tier 3 design
  • NVIDIA Enterprise GPUs
We provide GPUs from SBA Edge, an 80,000 square foot data center originally built for the United States government. It's equipped with industry-leading features and is able to withstand 90-mph winds. We have a multi-homed internet blend from SBA Edge's Chicago location, and we have the largest set of GPUs available from here.
Deployed GPUs:
  • GeForce GTX/RTX (On request with sufficient scale)
  • A100 SXM
  • A100 PCIE
  • V100 SXM
  • V100 PCIE
  • RTX A40
  • RTX A6000
  • RTX A5000
  • RTX A4000
  • Quadro RTX 4000
  • Quadro RTX 5000
iPerf Speedtest: ord1.speedtest.tensordock.com

Las Vegas DC1, Switch (las1)

  • On-demand instances, reservable instances
  • "Tier 5" design
  • NVIDIA Enterprise GPUs
We provide GPUs from Switch's Las Vegas campus. Our power source is 100% renewable. It's less than 10 ms to Los Angeles and is the perfect location for rendering workloads by VFX studios.
Deployed GPUs:
  • RTX A40
  • RTX A5000
  • RTX A4000
  • Quadro RTX 4000
  • Quadro RTX 5000
iPerf Speedtest: las1.speedtest.tensordock.com

Singapore DC1, Ascenix (sin1)

  • Reservable instances
  • NVIDIA GeForce GPU, NVIDIA Enterprise GPUs
Located at Ascenix with a premium IP blend, customers can run workloads from throughout Asia with low latencies. We operate this location in partnership with a local cloud gaming company with shared hardware ownership.
Deployed GPUs:
  • GeForce GTX/RTX (1070)
  • NVIDIA A-series
  • NVIDIA Quadro RTX

Singapore DC2, Equinix SG1 (sin2)

  • Reservable instances
  • NVIDIA GeForce GPU, NVIDIA Enterprise GPUs
Located at Equinix SG1 with a premium IP blend, customers can run workloads from throughout Asia with low latencies. We operate this location in partnership with a local cloud gaming company with shared hardware ownership.
Deployed GPUs:
  • GeForce GTX/RTX (1070)
  • NVIDIA A-series
  • NVIDIA Quadro RTX

Other Locations

These additional locations are used for our customers with specialized needs that aren't covered by our GPU hosting service. These are mostly used for our consulting and enterprise customers. Become one by contacting us.
  • TOR: Free exit nodes that we sponsor for the public good
  • VPN: Paid VPN services
  • Storage: File storage
  • Web: Web hosting
  • Compute: CPU-focused instances
  • Docker: Docker-enabled for both compute and storage workloads

Boston, United States — Self-Hosted (bos2)

  • VPN, Storage, Docker, Web, Compute

Chicago, United States — Digital Realty

  • VPN, Storage, Docker, Web

Los Angeles, United States — CoreSite

  • VPN, Storage, Docker, Web

Los Angeles, United States — Psychz Networks

  • VPN, Storage, Docker, Web, Compute
Note: We are phasing out Los Angeles-based compute services as of April 2022. Existing compute services will be migrated onto our own new hardware at bos2. Other Los Angeles-based services will remain.

Miami, United States — CoreSite

  • VPN, Storage, Docker, Web
Note: We are phasing out all Miami-based services as of October 2022. Existing services will be migrated to our clusters in OVH BHS.

Montreal, Canada — OVH BHS

  • VPN, Storage, Docker, Web, Compute

Montreal, Canada — eStruxture

  • VPN, Storage, Docker, Web

London, England — Digital Realty Meridian Gate

  • VPN, Storage, Docker, Web

Stockholm, Sweden — GleSYS Stockholm

  • VPN, Storage, Docker, Web, Compute
Note: We are phasing out Stockholm-based compute services as of April 2022. Existing compute services will be migrated onto our own new hardware at bos2. Stockholm-based storage services will remain.

Amsterdam, Netherlands — Equinix AM5

  • VPN, Storage, Docker, Web

Sandefjord, Norway — TerraHost

  • TOR, VPN, Storage, Docker, Web