i will not buy an rtx6000 i will not buy an rtx6000 i will not buy an rtx6000 i will not buy an rtx6000 i will not buy an rtx6000 i will not buy an rtx6000 i will not buy an rtx6000 i will not buy an rtx6000 i will not buy an rtx6000 i will not buy an rtx6000 i will not buy an rtx6000 i will not buy
September 28, 2026 at 6:39 AM
i will not buy an rtx6000 i will not buy an rtx6000 i will not buy an rtx6000 i will not buy an rtx6000 i will not buy an rtx6000 i will not buy an rtx6000 i will not buy an rtx6000 i will not buy an rtx6000 i will not buy an rtx6000 i will not buy an rtx6000 i will not buy an rtx6000 i will not buy
we have rtx6000 at home
Ternary Bonsai 2 (Qwen3.8-27B) at ~1200-1750tps prefill, ~100tps (peak ~400tps) decode on a 4080 super
Ternary Bonsai 2 (Qwen3.8-27B) at ~1200-1750tps prefill, ~100tps (peak ~400tps) decode on a 4080 super
September 19, 2026 at 12:26 AM
we have rtx6000 at home
Ternary Bonsai 2 (Qwen3.8-27B) at ~1200-1750tps prefill, ~100tps (peak ~400tps) decode on a 4080 super
Ternary Bonsai 2 (Qwen3.8-27B) at ~1200-1750tps prefill, ~100tps (peak ~400tps) decode on a 4080 super
Estimated hardware options to run this model locally:
$12k Mac Studio (2x 256GB or 1x 512GB)
$14k DGX GB-10 Spark (4x)
$14k AMD Strix cluster (4x)
$35k AMD Mi210 (8x)
$55k NVIDIA RTX6000 PRO (6x)
$12k Mac Studio (2x 256GB or 1x 512GB)
$14k DGX GB-10 Spark (4x)
$14k AMD Strix cluster (4x)
$35k AMD Mi210 (8x)
$55k NVIDIA RTX6000 PRO (6x)
April 14, 2026 at 4:06 PM
Estimated hardware options to run this model locally:
$12k Mac Studio (2x 256GB or 1x 512GB)
$14k DGX GB-10 Spark (4x)
$14k AMD Strix cluster (4x)
$35k AMD Mi210 (8x)
$55k NVIDIA RTX6000 PRO (6x)
$12k Mac Studio (2x 256GB or 1x 512GB)
$14k DGX GB-10 Spark (4x)
$14k AMD Strix cluster (4x)
$35k AMD Mi210 (8x)
$55k NVIDIA RTX6000 PRO (6x)
Congratulations on your new RTX6000
September 28, 2026 at 12:13 PM
Congratulations on your new RTX6000
PS5以上のスペックになりそうだからなぁ。
最低RTX4000、推奨がRTX6000当たりに合わせてくるのかな。
最低RTX4000、推奨がRTX6000当たりに合わせてくるのかな。
May 9, 2025 at 11:15 AM
PS5以上のスペックになりそうだからなぁ。
最低RTX4000、推奨がRTX6000当たりに合わせてくるのかな。
最低RTX4000、推奨がRTX6000当たりに合わせてくるのかな。
nate@RTX6000:~$ ollama show gemma3n:e4b | grep context
context length 32768
nate@RTX6000:~$ ollama show gemma3:270m | grep context
context length 32768
nate@RTX6000:~$ ollama show gemma3:27b | grep context
context length 131072
nate@RTX6000:~$
context length 32768
nate@RTX6000:~$ ollama show gemma3:270m | grep context
context length 32768
nate@RTX6000:~$ ollama show gemma3:27b | grep context
context length 131072
nate@RTX6000:~$
September 26, 2025 at 10:58 PM
nate@RTX6000:~$ ollama show gemma3n:e4b | grep context
context length 32768
nate@RTX6000:~$ ollama show gemma3:270m | grep context
context length 32768
nate@RTX6000:~$ ollama show gemma3:27b | grep context
context length 131072
nate@RTX6000:~$
context length 32768
nate@RTX6000:~$ ollama show gemma3:270m | grep context
context length 32768
nate@RTX6000:~$ ollama show gemma3:27b | grep context
context length 131072
nate@RTX6000:~$
More NVIDIA RTX Pro 6000 capacity is available NOW on Exoscale!
Built for teams working on AI inference, model testing, rendering, simulation and visual production on European cloud infrastructure.
www.exoscale.com/gpu/rtx6000/
#gpu #ai
Built for teams working on AI inference, model testing, rendering, simulation and visual production on European cloud infrastructure.
www.exoscale.com/gpu/rtx6000/
#gpu #ai
September 2, 2026 at 9:50 AM
More NVIDIA RTX Pro 6000 capacity is available NOW on Exoscale!
Built for teams working on AI inference, model testing, rendering, simulation and visual production on European cloud infrastructure.
www.exoscale.com/gpu/rtx6000/
#gpu #ai
Built for teams working on AI inference, model testing, rendering, simulation and visual production on European cloud infrastructure.
www.exoscale.com/gpu/rtx6000/
#gpu #ai
Check this #cloud offer from Linode 🚀
👉 1 * RTX6000 Ada 48 GB for $1.50/hour!!!
Book here: gpucompare.com/clusters/opp...
#cloud #ai #machinelearning #gpu
👉 1 * RTX6000 Ada 48 GB for $1.50/hour!!!
Book here: gpucompare.com/clusters/opp...
#cloud #ai #machinelearning #gpu
June 22, 2025 at 4:01 PM
Check this #cloud offer from Linode 🚀
👉 1 * RTX6000 Ada 48 GB for $1.50/hour!!!
Book here: gpucompare.com/clusters/opp...
#cloud #ai #machinelearning #gpu
👉 1 * RTX6000 Ada 48 GB for $1.50/hour!!!
Book here: gpucompare.com/clusters/opp...
#cloud #ai #machinelearning #gpu
Also, farms benefit way, WAY more people. The pro-AI people don't care outside of money.
Other thought:
How many chickens can each $10,000 RTX6000 Pro buy to feed us?
Other thought:
How many chickens can each $10,000 RTX6000 Pro buy to feed us?
September 6, 2026 at 7:56 PM
Also, farms benefit way, WAY more people. The pro-AI people don't care outside of money.
Other thought:
How many chickens can each $10,000 RTX6000 Pro buy to feed us?
Other thought:
How many chickens can each $10,000 RTX6000 Pro buy to feed us?
RTX6000番あるのかなと調べたら…あった
Quadroだった
Quadroだった
October 27, 2025 at 2:58 PM
RTX6000番あるのかなと調べたら…あった
Quadroだった
Quadroだった
The nerd in me:
"Sapphire only makes AMD GPU's. They mostly use Nvidia RTX6000 ADA and RTX5090's."
"Sapphire only makes AMD GPU's. They mostly use Nvidia RTX6000 ADA and RTX5090's."
February 14, 2026 at 8:12 PM
The nerd in me:
"Sapphire only makes AMD GPU's. They mostly use Nvidia RTX6000 ADA and RTX5090's."
"Sapphire only makes AMD GPU's. They mostly use Nvidia RTX6000 ADA and RTX5090's."
Like MultiFlow this does struct+seq, with struct similar to FoldFlow-SFM, and with seq per Meta's Discrete Flow Matching. We trained this for a week on one RTX6000 Ada, expecting sketchy helices, and maybe the odd beta sheet pairing if we were lucky. Instead, we got these:
June 13, 2025 at 5:06 PM
Like MultiFlow this does struct+seq, with struct similar to FoldFlow-SFM, and with seq per Meta's Discrete Flow Matching. We trained this for a week on one RTX6000 Ada, expecting sketchy helices, and maybe the odd beta sheet pairing if we were lucky. Instead, we got these:
The GPU die prices are also artificially inflated for Nvidia because any die they use for a RTX 5xxx is a lost $10k sale on a RTX6000 which will sell out just as instantly. And Nvidia is part of the cartel at this point, they are colluding with micron and OpenAI to price fix.
January 1, 2026 at 1:41 AM
The GPU die prices are also artificially inflated for Nvidia because any die they use for a RTX 5xxx is a lost $10k sale on a RTX6000 which will sell out just as instantly. And Nvidia is part of the cartel at this point, they are colluding with micron and OpenAI to price fix.
yeah it's not HBM in this case but on my the RTX6000 it... actually isn't HBM either, looking at the specs. i assumed it was.
August 17, 2026 at 6:03 PM
yeah it's not HBM in this case but on my the RTX6000 it... actually isn't HBM either, looking at the specs. i assumed it was.
you're right.. & CUDA is way more polished than MLX right now.. but M5 Ultra Studio's can compete with RTX6000
tok/watt is better on Macs.. and software is easier these days.. MLX will improve
and IMO: nvidia is ripe for competitors.. and I think Apple is better positioned than AMD for consumers
tok/watt is better on Macs.. and software is easier these days.. MLX will improve
and IMO: nvidia is ripe for competitors.. and I think Apple is better positioned than AMD for consumers
September 4, 2026 at 1:31 PM
you're right.. & CUDA is way more polished than MLX right now.. but M5 Ultra Studio's can compete with RTX6000
tok/watt is better on Macs.. and software is easier these days.. MLX will improve
and IMO: nvidia is ripe for competitors.. and I think Apple is better positioned than AMD for consumers
tok/watt is better on Macs.. and software is easier these days.. MLX will improve
and IMO: nvidia is ripe for competitors.. and I think Apple is better positioned than AMD for consumers
I really like my Asus ZenBook S14. YMMV.
I don't develop on it directly. Everything goes through ssh/tailscale to my big home machine, 9995wx/768gb/96gb rtx6000 pro Blackwell. I can't lug that compute with me.
Spending a lot of money specifically on a fast *laptop* seems like a bad idea to me.
I don't develop on it directly. Everything goes through ssh/tailscale to my big home machine, 9995wx/768gb/96gb rtx6000 pro Blackwell. I can't lug that compute with me.
Spending a lot of money specifically on a fast *laptop* seems like a bad idea to me.
July 30, 2026 at 8:16 AM
I really like my Asus ZenBook S14. YMMV.
I don't develop on it directly. Everything goes through ssh/tailscale to my big home machine, 9995wx/768gb/96gb rtx6000 pro Blackwell. I can't lug that compute with me.
Spending a lot of money specifically on a fast *laptop* seems like a bad idea to me.
I don't develop on it directly. Everything goes through ssh/tailscale to my big home machine, 9995wx/768gb/96gb rtx6000 pro Blackwell. I can't lug that compute with me.
Spending a lot of money specifically on a fast *laptop* seems like a bad idea to me.
Check this #Cloud offer from Linode 🚀
👉 1 * RTX6000 Ada 48 GB for $1.50/hour!!
Book here:https://gpucompare.com/providers/linode
#cloud #ai #machinelearning #gpu
👉 1 * RTX6000 Ada 48 GB for $1.50/hour!!
Book here:https://gpucompare.com/providers/linode
#cloud #ai #machinelearning #gpu
September 14, 2025 at 4:00 PM
Check this #Cloud offer from Linode 🚀
👉 1 * RTX6000 Ada 48 GB for $1.50/hour!!
Book here:https://gpucompare.com/providers/linode
#cloud #ai #machinelearning #gpu
👉 1 * RTX6000 Ada 48 GB for $1.50/hour!!
Book here:https://gpucompare.com/providers/linode
#cloud #ai #machinelearning #gpu
UpCloud expands its GPU lineup with new NVIDIA RTX Pro 6000 servers. With 96 GB of GPU memory, they're built for visualization, 3D rendering, and AI inference — billed only while powered on. 🚀 #GPU #AI #NVIDIA #RTX6000 #CloudComputing
New GPU Servers: NVIDIA RTX Pro 6000
NVIDIA RTX Pro 6000 GPU Servers added to the lineup for graphics, visualization, 3D rendering, and inference-heavy workloads.
upcloud.com
September 16, 2026 at 4:00 PM
UpCloud expands its GPU lineup with new NVIDIA RTX Pro 6000 servers. With 96 GB of GPU memory, they're built for visualization, 3D rendering, and AI inference — billed only while powered on. 🚀 #GPU #AI #NVIDIA #RTX6000 #CloudComputing
「今後出るRTX6000シリーズも6,000円台にしてほしい」と寝言を言っていたらRTX600090が発売された回。
February 12, 2026 at 2:26 PM
「今後出るRTX6000シリーズも6,000円台にしてほしい」と寝言を言っていたらRTX600090が発売された回。
The results on both NVIDIA GeForce RTX 4090 and NVIDIA RTX6000 show throughput improvements of up to 1.80x for data generation, 1.85x for RL training, and 1.23x for VLA-based RL.
May 5, 2026 at 6:40 AM
The results on both NVIDIA GeForce RTX 4090 and NVIDIA RTX6000 show throughput improvements of up to 1.80x for data generation, 1.85x for RL training, and 1.23x for VLA-based RL.
yooooo this build looks amazing! youtu.be/E91g41Irva0?...
「BRO」4K Water Cooled PC Build 🔪 1000D Overkill NVIDIA RTX6000 Ada Generation 48GB x5 83KUSD#pcbuild
YouTube video by BRO COOLING
youtu.be
November 8, 2024 at 11:50 PM
yooooo this build looks amazing! youtu.be/E91g41Irva0?...
NVIDIA pourrait suspendre ses GPU gaming en 2026 et repousser la RTX 6000 à 2028, impactée par la pénurie de mémoire et la priorité donnée à l’IA.
#NVIDIA #RTX6000 #PCGaming #IA
www.justjeuxvideo.com/nvidia-boule...
#NVIDIA #RTX6000 #PCGaming #IA
www.justjeuxvideo.com/nvidia-boule...
NVIDIA bouleverse le gaming PC, la RTX 6000 repoussée à 2028 et l’IA prend le dessus
NVIDIA pourrait suspendre ses GPU gaming en 2026 et repousser la RTX 6000 à 2028, impactée par la pénurie de mémoire et la priorité donnée à l’IA.
www.justjeuxvideo.com
February 8, 2026 at 6:30 AM
NVIDIA pourrait suspendre ses GPU gaming en 2026 et repousser la RTX 6000 à 2028, impactée par la pénurie de mémoire et la priorité donnée à l’IA.
#NVIDIA #RTX6000 #PCGaming #IA
www.justjeuxvideo.com/nvidia-boule...
#NVIDIA #RTX6000 #PCGaming #IA
www.justjeuxvideo.com/nvidia-boule...
Example run:
Training via mirror self-play on 3× RTX6000 Ada (~12h) led to strong gains vs. google/gemini-2.0-flash-lite-001—both in-domain (SimpleTak) and OOD (KuhnPoker). Learned strategies generalize, despite never playing vs. Gemini during training.
Training via mirror self-play on 3× RTX6000 Ada (~12h) led to strong gains vs. google/gemini-2.0-flash-lite-001—both in-domain (SimpleTak) and OOD (KuhnPoker). Learned strategies generalize, despite never playing vs. Gemini during training.
June 24, 2025 at 4:45 PM
Example run:
Training via mirror self-play on 3× RTX6000 Ada (~12h) led to strong gains vs. google/gemini-2.0-flash-lite-001—both in-domain (SimpleTak) and OOD (KuhnPoker). Learned strategies generalize, despite never playing vs. Gemini during training.
Training via mirror self-play on 3× RTX6000 Ada (~12h) led to strong gains vs. google/gemini-2.0-flash-lite-001—both in-domain (SimpleTak) and OOD (KuhnPoker). Learned strategies generalize, despite never playing vs. Gemini during training.