Qubitium

Building GPT-QModel, ModelCloudAI. Contributor to stuff you are probably using.

[Qubitium @qubitium]

🚀 GPT-QModel v7.0.0 is here!

🔥 Huawei Ascend NPU support
⚡ JIT compilation for CUDA/ROCm kernels on first use
✅ Easier Pip/UV installs with no --no-build-isolation needed

🧠 Models Added: ⚡ GLM 5 / 5.1 👁️ GLM OCR 🎙️ GLM ASR 💎 Gemma 3n 🦅 Falcon Mamba 🖼️ InternVL Chat

[Qubitium @qubitium]

@HotAisle Did you ever play with arm dc chips like ampere? Were they that bad?

[Qubitium @qubitium]

total direct to cpu, no plx, 176 pcie channels if you count the mcio x8 plus x4 connectors.

[TechPowerUp @TechPowerUp]

NVIDIA "Vera" CPU Benchmarked: Beating Intel Xeon and AMD EPYC in Select Workloads https://t.co/SwzxnNtmxX

[Qubitium @qubitium]

tdp is max 600w per cpu based on the 16p power connector.

[Qubitium @qubitium]

@anemll You got those skus from grok? 💀 I thought they were real. lol

[Qubitium @qubitium]

Us tax payers not only get shafted once, but twice. First in my ins premiums, and even worse, another is waiting at the ER. I am leaving California.

[T Wolf 🌁 @Twolfrecovery]

California provided full scope Medi-Cal health coverage to 2 million undocumented immigrants between 2020-2024, free. How much do you pay for your health insurance?

[Qubitium @qubitium]

My posts are for the xeon 4th gen max cpus. Not sure if they have fixed this for the skus you mentioned which appears to be newer. There are also actual hw bugs that regessed l3 cache timing and cut bandwidth by 1/4 when ht is enabled. it was not a failure, it was an engieering disaster at every level

[Qubitium @qubitium]

🪣 Bucket list check: In the rain, in the dark, I helped two AC guys pull a 200 lb, 4-ton / 48k BTU/h condenser 10 feet up via a carabiner pulley in a disgusting back-kitchen alley.

Not my plan for today.

But lab servers must be cooled.

My motto: by any means necessary.

[Qubitium @qubitium]

If I am CIA and I have to either control the narrative of AI or cripple the competing nations ability to react, she would be on my top list of recruits.

[Karen Hao @_KarenHao]

On the one-year anniversary of EMPIRE OF AI, I am so, so excited to announce The AI Resist List, a new project that documents examples of resistance to the AI empires around the world.

https://t.co/mJqk5FxkFY

[Qubitium @qubitium]

@sheriyuo LOL

[Qubitium @qubitium]

Humidity could be it!

The label core appears to be synthetic fabric coated with plastic/polymer so sn/model ids can be printed on. The polymer coating that gives you the plastic feel is the one degraded. I found more cables labels with same issue in this node. but some are in perfect condition.

If we assume all are original cables but only some (most) are degraded, all points to supplier and batch qa issue.

[Qubitium @qubitium]

Why oem server degrades after 6 years and you won't believe the reason why. 🤓

[Qubitium @qubitium]

The cable in question is a 15cm sas cable inside a small blade enclosure for a storage node.

[Qubitium @qubitium]

Old but good enough as zfs storage host/controller. Xeon 8173M

[Qubitium @qubitium]

CTranslate is a hidden gem btw.

[Jordi Mas @jordimash]

I’ve just released CTranslate2 4.7.2 (https://t.co/ffxZ4U5am2). This new version includes support for Gemma 4 31b dense model and includes several bug fixes and improvements.

[Qubitium @qubitium]

@sheriyuo Where was the announcement held? bj or sh?

[Qubitium @qubitium]

@shisoftgenius You have blowers front to back, why are you installing 2 ram fans? Does it actually help? From the looks of it, it might actually hinder the heat escaping.

[Qubitium @qubitium]

They did nuke H20's FP64 performance to 1 TFLOPS from.......34 TFLOPS. Pretty much banned. 🫶

[Luigi Cruz @luigifcruz]

Calling GPUs “nuclear bombs” is unserious policy theater. At this rate, they should advocate banning FP64 math too.

[Qubitium @qubitium]

@aleks_sharik 👀

[Qubitium @qubitium]

I absolutely agree the 1-8 node non-training compatible H100 instance pricing spike is likely manipulated by someone or entity and not actually used for ai.

[Brandon Carl @brandonjcarl]

Crazy price action in H200 cloud pricing – up 56% in 3 days.

What is unusual is that the H200 is suddenly trading higher than the B200, a superior GPU.

It’s not crazy to think that a fund could bid up supply in an illiquid tight market at a cost of $50K a day to engineer a short-term move in much more liquid stocks.

[Qubitium @qubitium]

@DeepIceValue No