Hacker News Reader

Comment by oblivuslimited | Hacker News Reader

Don't think you should manufacture engagement by asking people to comment in order to get free credits.

We considered creating a $25 promo code that anyone could use without contacting us, but the reason we chose this method is because even with a $1 promo code last time, we saw many fake accounts spamming it.

That said, we still have a $3 promo code that anyone can use without needing to comment here: HN_3. Also, if you prefer not to comment publicly, feel free to send your feedback to feedback@oblivus.com!

Right, but why not use GitHub accounts?

Well, because we currently don’t have GitHub integration on the platform, but that’s a good point! We’ll definitely consider adding SSO in the future. Thank you!

This is brilliant, quite cheaper than my AWS. Can I get the credits please? bp2fe1b2a8c422cf09b2a375a

Hello spk_, the necessary credits have successfully been allocated to your account. We’d love to hear your feedback—share your thoughts and you can earn another $15 in credits!

Here are some feedback/questions: 1. I was logged in via ssh and when I tried the web shell it kicked me and the machine went unreachable. 2. How can I restrict who can access my machine? There is no allowed IPs section. 3. Can I upload my ssh-key somewhere and use that as a default whenever I launch an instance? 4. Why do I even have images section when I can't take a snapshot and keep it as a image? This would be very handy so that I don't need to repeat the basic setup of my instances. 5. Can I have a auto shutdown based on time? I don't want to accidentally leave my machine running and rake up a huge cost.

Hope this helps.

There's something odd about the pricing in the invoice, I assume you are doing per second billing? I ran a test instance for a few mins and this is what I see in the invoice:

CPU Usage - Intel Xeon v3 1.5333 hours $0.044467 Storage Usage - NVMe 3.3333 hours $0.000333 Public IP Address 0.4000 hours $0.002200

The price seems to be ok but the units are way off.

Hey there,

1. That’s interesting and it shouldn’t happen. I just tested it myself and was able to connect via both VNC and Terminal at the same time. I’ll check with the team to see if there’s an issue. 2. We don’t impose any physical firewalls on the VMs, but you can set up a virtual firewall of your choice and limit IP access. 3. We try to avoid storing any critical customer information. Public key authentication is possible, but you’ll need to share your key each time on the deployment page. We’ll explore ways to save the key securely! 4. It is possible to create images. This was automated before, but we recently had to switch to a manual process due to changes in our hypervisor software. If you reach out to support, they can create images from your VM and attach them to your billing profile, usually within a day. 5. This is a feature we don’t currently have, but it’s definitely something we’ll consider adding to our roadmap!

Regarding pricing: Every resource is billed on a per-minute basis. For instance, if you have 100GB of storage, the system will record 100 minutes of NVMe usage every minute. The same applies to vCPUs and GPUs. I hope that clarifies things, but if you need further details please feel free to ask.

Thanks a lot for the feedback, really appreciate it! We’ve gone ahead and applied the remaining credits to your account.

Here is my billing profile: bp655f42f5e90faba9cbd5805. Looking forward to testing it!

Hello KNorbrendy, we have successfully allocated the necessary credits to your account. We’d love to hear your feedback—share your thoughts and you can earn another $15 in credits!

The UI looks great and is very user-friendly. I liked that you can see the real-time availability.

I spun up a few instances, and all were up and running within 1-3 minutes. The speedtests were around 7Gbps, which is more than enough for my case. The CPUs in the machines I got were either AMD 7xxx or 9xxx models. I also really like that ingress and egress are free of charge, this is a big issue we face with AWS. Pricing was straightforward as well, and I liked seeing real-time updates based on server status.

One suggestion would be to add custom balance alerts. It seems like this feature isn’t available yet, but I’d love the option to set alerts when my balance drops below a certain threshold.

Hello KNorbrendy, thank you for your feedback! We’ve activated the necessary credits on your account. Just a quick reminder, if you explore the API, we can add an additional $5 credit!

This is super! Best of luck ! BILLING PROFILE: bp7cc3b60e5ce167e99b1ce77

Hello dwaiba, the necessary credits have successfully been allocated to your account. We’d love to hear your feedback—share your thoughts and you can earn another $15 in credits!

Thanks. Looks good but do the GPUs like "1x H100 80GB NVLINK" support Infiniband?. Are these HGX Modules? (Since standard HGX comes in pairs of 4 or 8). Also, when will H200 be available? Great Stuff !

Thanks for the feedback! NVLink configurations are available only in 8x, it’s not possible to deploy 1x. These are PCIe modules connected via NVLink bridges, and they are not HGX.

We do plan to expand our fleet with B200s, but H200s are not currently on the roadmap for on-demand.

We’ve just activated the rest of your credits—thank you again for your feedback!

1. I tried with L40 for inference in MON1 with BentoML [@bentoml.service(resources={"gpu": 1}) class MyService: def __init__(self): import torch self.model = torch.load('model.pth').to('cuda:0')]

and

2. RTX A4000 for training using pytorch training script [ device = "cuda" if torch.cuda.is_available() else "cpu" device] in OSL1 with sshkeys for root.

Old CUDA 12.2 is provided with Mainline Ubuntu LTS 22.04 LTS as Latest Ubuntu.

It is unstable (mostly) with driver upgrades as Device drops out though works intermittently !!!

Start Scripts for pyenv and CUDA: #!/bin/bash

DEBIAN_FRONTEND=noninteractive apt-get -y install wget zip unzip git jq

DEBIAN_FRONTEND=noninteractive apt-get install -y python3-pip

# Required for pyenv make DEBIAN_FRONTEND=noninteractive apt-get install -y build-essential zlib1g-dev libffi-dev libssl-dev libbz2-dev libreadline-dev libsqlite3-dev liblzma-dev libncurses-dev tk-dev DEBIAN_FRONTEND=noninteractive apt install -y --reinstall gcc-12 ln -s -f /usr/bin/gcc-12 /usr/bin/gcc # Remove existing nvidia drivers and install cuda 12.6.1 DEBIAN_FRONTEND=noninteractive apt-get -y remove --purge *nvidia* DEBIAN_FRONTEND=noninteractive apt-get -y remove --purge *cuda* DEBIAN_FRONTEND=noninteractive apt-get -y remove --purge *nvrtc* DEBIAN_FRONTEND=noninteractive apt-get -y autoremove --purge wget https://developer.download.nvidia.com/compute/cuda/12.6.1/lo... && sh cuda_12.6.1_560.35.03_linux.run --silent --override DEBIAN_FRONTEND=noninteractive apt-get -y update DEBIAN_FRONTEND=noninteractive apt-get -y upgrade DEBIAN_FRONTEND=noninteractive reboot

Thank you for the feedback! We’ve actually just updated the default Ubuntu OS to 550.90.12 and CUDA to 12.4 today. Would you mind checking it again?

Competition is always good! best of luck! bp762a7e430ca313d22506624

Hello rkwasny,the necessary credits have successfully been allocated to your account. We’d love to hear your feedback—share your thoughts and you can earn another $15 in credits!

Can I get the credit? bp5c642519d8152c4e4f019f6

Hello Bookidex,the necessary credits have successfully been allocated to your account. We’d love to hear your feedback—share your thoughts and you can earn another $15 in credits!

Very cool.I wish I had the account earlier

Hey there, you can still use the HN_3 code to get some credits. Feel free to share your feedback, and we’ll see what we can do for you!

Reply on news.ycombinator.com