https://xitter.cf/t.co/dfIHZ8g6ND: Affordable, top-tier cloud hosting for AI. Empowering innovators with 80GB A100 GPUs. Rent by hour or long-term. We're here for you.
Weehawken, NJ
Joined June 2023
- Tweets433
- Following35
- Followers404
- Likes689
Pinned Tweet
🚨Hey everyone, I've got about 10 A100 servers available for rent.🚨
Each one comes with:
8× A100 80GB GPUs
Dual 64-core AMD Milan 7763 CPUs
1TB RAM
8TB NVMe storage
Uncapped internet
Rent is $10,000 per month per server.
If you're interested or want to know more, just send me a DM. #AI #MachineLearning #GPU #LimitedOffer
Redmond Hosting retweeted
Pennywise - Dance With Your Demons (Teaser Trailer)
I've spent the last 2 months on a new music video and excited to finally announce it to you guys!
#pennywise #Dallasxy #welcometoderry #musicvideo #trailer
Redmond Hosting retweeted
Cinematic destruction!
I'm excited to share my Explosion LoRA for @Alibaba_Wan 2.2 -I2V that transforms static images into dramatic explosion sequences with realistic physics and effects!
Special thanks to @RedmondAI for their generous support in making this happen. Check it out. links coming in the next tweet!
Redmond Hosting retweeted
See the world in a whole new light, thermal light!
Im sharing today my Spectramatic LoRA for @Alibaba_Wan 2.2 that transforms ordinary scenes into stunning thermal vision videos. Watch as heat signatures paint the world in vibrant colors!
Special thanks to @RedmondAI for their generous support in making this happen.
Link next tweet.
I'm really interested to see what happens in 2 years when these mega DCs are now filled with outdated silicon. Surely Nvidia must know that 200k GPUs aren't going to be replaced every year.
How current will these DCs stay? What will happen to the GPU market when hundreds of thousands of $20k GPUs are offloaded?
Can LLMs vibe code FPGAs yet? Once this can happen reliably it could be a huge push to evolve FPGA to be more powerful and cheaper. Each app could have its own unique hardware accelerator .
Literally 3 out of 4 posts on my wall are political now for some reason. I say I am not interested in all of them, but they still keep coming.
o3 is really interesting the way it works.
Unlike today’s LLMs, which typically generate answers in a single pass, o3 explores a huge variety of possible solution “programs” on the fly.
Think of it like AlphaZero for thinking: it searches a massive space of reasoning steps, tests them out, and uses a learned evaluator to zero in on the most promising paths.
The result? A model that can tackle truly novel challenges and combine knowledge in ways older models can’t—though it comes at a high computational cost. It’s a glimpse at how the next wave of AI might genuinely “think” through complex, unfamiliar problems.
We are looking for a developer to help us fine-tune a modern (mochi, cogvideox) video generation model on inpainting. For the right experience we are open to any salary.
Essentially we want to recreate the inpainting examples of this model lumiere-video.github.io/#sec…
If you have the experience and you are interested please DM us!
Redmond Hosting retweeted
🎬 The official trailer for The Kharzov Effect is here! 🎬
🔗 youtu.be/8Pc4kUM-59A
Discover the dark secrets of our 80s horror puzzle game and get a glimpse of the mysteries waiting for you. 🕵️♂️
But there’s more... Hidden within the trailer is a puzzle. Solve it to unlock an exclusive code—redeemable for future in-game rewards! 🧠
Are you sharp enough to crack the code? 👁️
#horrorgame #HorrorGames #Steam #Kickstarter
I decided to create an OFX plugin. Fed some python code and 700+ pages of OFX documentation to o1 and it spits out a plugin!
Last step! Now I post the code, plus the 700 pages of OFX documentation and add this
"Take a look at this OFX plugin and then read through the OFX documentation. Do you see anything in the plugin that wouldn't compile correctly based on the documentation? Do you see any errors in formatting, syntax, or how methods are used?"
Do that a couple times and Voila! Now it is time for me to actually compile and test :)
Even with long context windows the LLM can start going in circles, thats why I found it is best to always start with a blank slate. If I wasn't so lazy I would write a script that did all this for me and I just come back in half an hour and pay the $10 in o1 calls.