GPU Servers for LLM - Any deals?

hostinghunter

Do you know any providers? I see it's costly with most providers. Is there a racknerd of GPU Servers? Everything else is super expensive.

cainyxues

Crunchbits has GPU servers right? @sh97

sh97

@cainyxues said:
Crunchbits have GPU servers right? @sh97

Thanks for the mention!

We have some GPU VDS which might work for you.
https://crunchbits.com/gpu/cloud

hostinghunter

@sh97 said:

@cainyxues said:
Crunchbits have GPU servers right? @sh97

Thanks for the mention!

We have some GPU VDS which might work for you.
https://crunchbits.com/gpu/cloud

Any limitations? Looking to deploy multiple LLMs.

fiberstate

Depending on how much GPU you need, we have a few options.

We have 4090 GPU add-ons for any of our Ryzen (Ryzen 7 5700G, or 9 9750X/9950X) lineup of servers.

Our Ryzen 7 5700G servers come with a iGPU that can be used, you can also add a dedicated Nvidia 4090 GPU

We have 4090's in stock and ready for deployment and yes you can deploy LLM's on our network.

---

AMD Ryzen 7 5700G @ 3.8 GHz (8 Cores/16 Threads)

128GB Memory RAM (double ram promo)
1TB NVMe Samsung Evo Plus
1 IPv4 Address Included IPv4
/64 IPv6 Addresses available upon request
1 Gbps Unlimited Network Port
Included (reboot, install, console) VPN IPMI
Salt Lake City Data Center
$49.95/mo 

👉 Order Now

---

BasToTheMax

TensorDock maybe?

sh97

@hostinghunter said:

@sh97 said:

@cainyxues said:
Crunchbits have GPU servers right? @sh97

Thanks for the mention!

We have some GPU VDS which might work for you.
https://crunchbits.com/gpu/cloud

Any limitations? Looking to deploy multiple LLMs.

Nothing, it's dedicated resources
If you see something out of stock, but it interests you, shoot me a DM, will add an internal order if we have the resources.

vpsam

@sh97 said:

@cainyxues said:
Crunchbits have GPU servers right? @sh97

Thanks for the mention!

We have some GPU VDS which might work for you.
https://crunchbits.com/gpu/cloud

I'd suggest maybe adding in the VRAM of the different GPU models on that page.

johndeo983

shadow.tech A4500 55 euro/m

vpsam

@johndeo983 said:
shadow.tech A4500 55 euro/m

This is not a server though (unless thing's have changed), but a remote PC which shuts down automatically after the keyboard/mouse has been inactive for a while.

crunchbits

@vpsam said:

@sh97 said:

@cainyxues said:
Crunchbits have GPU servers right? @sh97

Thanks for the mention!

We have some GPU VDS which might work for you.
https://crunchbits.com/gpu/cloud

I'd suggest maybe adding in the VRAM of the different GPU models on that page.

Good note, will add to tasks for next update. Thank you

hostinghunter

These are good options.

I need 4 servers that can handle the following needs

1. Reasoning like ChatGPT, DeepSeek, Claude, etc.
2. Image creation
3. Audio generation including tone, voice, gender, and language options.
4. Video generation, including integration audio from above. The video supports avatars based on gender, avatar, lips, voice match, and localization. Add: Custom Avatar (Uses Personalised Avatars).

Is it possible to deploy multiple models on a single server? Any restrictions?

Any Suggestions?

generaL

@hostinghunter said:
These are good options.

I need 4 servers that can handle the following needs

1. Reasoning like ChatGPT, DeepSeek, Claude, etc.
2. Image creation
3. Audio generation including tone, voice, gender, and language options.
4. Video generation, including integration audio from above. The video supports avatars based on gender, avatar, lips, voice match, and localization. Add: Custom Avatar (Uses Personalised Avatars).

Is it possible to deploy multiple models on a single server? Any restrictions?

Any Suggestions?

It completely depends on the LLM you are going to use.
The feature you mentioned requires multiple high end gpu to get desired output.

May i know Which llm you want to run?
You probably can find reviews about which gpu performs better for your desired LLM.
Also if you want to publicly launch your AI chat bot similar to pou or other platform, using the api might be cheaper option.
Check openrouter for cheap api.
Good luck

SlowDD

@hostinghunter said:
These are good options.

I need 4 servers that can handle the following needs

1. Reasoning like ChatGPT, DeepSeek, Claude, etc.
2. Image creation
3. Audio generation including tone, voice, gender, and language options.
4. Video generation, including integration audio from above. The video supports avatars based on gender, avatar, lips, voice match, and localization. Add: Custom Avatar (Uses Personalised Avatars).

Is it possible to deploy multiple models on a single server? Any restrictions?

Any Suggestions?

It's hard to give suggestions for such a broad topic without a lot of background information. Step 4 is the most crucial one. I if it's just a hobby, you can use svd for it, else LTX is a good idea.

letstayfree

I'd like have cheaper one for testing. But no idea about one like racknerd