4 ms·

FYI Google's K80 ($0.7, pay-per-minute) are cheaper than AWS's K80 ($0.9/h, pay-per-hour), as well as Azure's K80 ($1.08/h, pay-per-hour).

by sheerun 10y ago

boulos 10y ago
Yep! And for distributed training, the per-minute billing is incredibly important. But just a correction: Azure also does per-minute billing (following suit from GCE way back in the day).
minimaxir 10y ago
Google's $0.70 rate is in addition to the normal instance cost, so math is more complicated if you don't have a machine already. (Alhough, a low-end high memory machine is $0.126/hr, so the sum is still lower than Amazon)
- boulos 10y ago
  Yes, but you can connect any number of GPUs to an n1-standard-1, so feel free to add $.05/hr :).
  - gwern 10y ago
    Is the n1-standard-1 instance fast enough to not be the bottleneck for the K80?
    - boulos 10y ago
      It really depends on what you're doing. If you are doing large transfers over PCIe back and forth, not really. But lots of things work just fine. The bigger challenge for large ML models is the memory you'd need to back it. But with GCE you can happily do a custom machine with up to 6.5 GB per vCPU, just so you can fit the output ;).
  - NegatioN 10y ago
    Slightly off-topic, but I don't know where to get help with this. I see one of our projects has gotten a quota of 16 GPUs in asia-east1, us-central1, us-east1. However we seem to have been allocated nothing in europe-west1. Is this an error, or simply something I have to manually ask for somewhere? "Quota 'NVIDIA_K80_GPUS' exceeded. Limit: 0.0" :( I'm so psyched to finally be able to use GPUs on GCE instead of AWS, so any help would be appreciated.
ktta 10y ago
What about the cost reduction that usually happens with extended use called sustained discounts? Does anyone have that info?
- boulos 10y ago
  Good question! We do not currently apply sustained-use discounts to GPUs, but we may do so in the future (we need to gather data first on usage, to understand if people will be running 24x7) [1]: > You cannot attach GPUs to preemptible instances. GPUs do not receive sustained use discounts. [1] https://cloud.google.com/compute/pricing#gpus https://cloud.google.com/compute/pricing#gpus

FYI Google's K80 ($0.7, pay-per-minute) are cheaper than AWS's K80 ($0.9/h, pay-per-hour), as well as Azure's K80 ($1.08/h, pay-per-hour).

by sheerun 10y ago

FYI Google's K80 ($0.7, pay-per-minute) are cheaper than AWS's K80 ($0.9/h, pay-per-hour), as well as Azure's K80 ($1.08/h, pay-per-hour).

boulos 10y ago

Yep! And for distributed training, the per-minute billing is incredibly important. But just a correction: Azure also does per-minute billing (following suit from GCE way back in the day).

minimaxir 10y ago

Google's $0.70 rate is in addition to the normal instance cost, so math is more complicated if you don't have a machine already. (Alhough, a low-end high memory machine is $0.126/hr, so the sum is still lower than Amazon)

boulos 10y ago
Yes, but you can connect any number of GPUs to an n1-standard-1, so feel free to add $.05/hr :).
- gwern 10y ago
  Is the n1-standard-1 instance fast enough to not be the bottleneck for the K80?
  - boulos 10y ago
    It really depends on what you're doing. If you are doing large transfers over PCIe back and forth, not really. But lots of things work just fine. The bigger challenge for large ML models is the memory you'd need to back it. But with GCE you can happily do a custom machine with up to 6.5 GB per vCPU, just so you can fit the output ;).
- NegatioN 10y ago
  Slightly off-topic, but I don't know where to get help with this. I see one of our projects has gotten a quota of 16 GPUs in asia-east1, us-central1, us-east1. However we seem to have been allocated nothing in europe-west1. Is this an error, or simply something I have to manually ask for somewhere? "Quota 'NVIDIA_K80_GPUS' exceeded. Limit: 0.0" :( I'm so psyched to finally be able to use GPUs on GCE instead of AWS, so any help would be appreciated.

ktta 10y ago

What about the cost reduction that usually happens with extended use called sustained discounts? Does anyone have that info?

boulos 10y ago
Good question! We do not currently apply sustained-use discounts to GPUs, but we may do so in the future (we need to gather data first on usage, to understand if people will be running 24x7) [1]: > You cannot attach GPUs to preemptible instances. GPUs do not receive sustained use discounts. [1] https://cloud.google.com/compute/pricing#gpus https://cloud.google.com/compute/pricing#gpus