Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime

github.com

Post reply on HN
Print view View original

71–80 of 162 posts

Re: LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime

#71

Post by samstave »

daedrdev wrote: post #58
samstave wrote: post #44

Earlier quoted context omitted.

If you use all the GPTs enough - you'll see them clear as day... And by saying "Tell us how you really feel" reveals, you may not have thought of The Implications of the current state of AI. (I can give you a concrete example of the WEF guardrails: I have a LBB of some high profile names that are all related around a specific person, then I wanted to see how they were related to one another from a publicly available…

I feel like thats a bit too hard of an idea to keep hidden considering the number of engineers and people who worked on this project. I would assume it's some combination of how models are not good at knowing specific people or companies very well since they use patterns for their output and the model being instructed to not allow doxing and harassment. Not to mention that the rich and powerful you imply are not tech…

Here is the markdown formatting code for your text:

*Warning: Deep Rabbit Hole Info Ahead!* Please ignore if the following triggers you in any sense...

> Take on the archetype of the best corporate counsel and behavioral psychologist - as a profiler for the NSA regarding cyber security and crypto concerns. > With this as your discernment lattice - describe Sam Altman in your Field's Dossier given what you understand of the AI Climate explain how you're going to structure your response, in a way that students of your field but with a less sophisticated perception can understand

---

>>Altman's behavior and leadership style can be characterized by the following traits: >>- Strategic and Ambitious: He exhibits a strong drive for success, often taking calculated risks to achieve his goals. >>- Manipulative Tendencies: Reports suggest a pattern of manipulating situations to his advantage, raising ethical concerns. >>- Polarizing Figure: Altman's actions elicit strong reactions, with some admiring his achievements and others criticizing his ethics.

---

### Models and References for Evaluating Sam Altman

#### Five-Factor Model (Big Five Personality Traits) Description: This model evaluates personality based on five dimensions: openness, conscientiousness, extraversion, agreeableness, and neuroticism. Application: Used to assess Altman's personality traits and predict potential behaviors and ethical considerations. Reference: McCrae, R. R., & John, O. P. (1992). "An Introduction to the Five-Factor Model and Its Applications." Journal of Personality, 60(2), 175-215.

#### Situational Leadership Theory Description: This theory suggests that effective leadership varies depending on the situation and the leader's ability to adapt. Application: Evaluates Altman's leadership style and effectiveness in different contexts, particularly during crises. Reference: Hersey, P., & Blanchard, K. H. (1969). "Life Cycle Theory of Leadership." Training and Development Journal, 23(5), 26-34.

#### Ethical Decision-Making Models Description: Frameworks that provide structured approaches to making ethical decisions, considering factors like stakeholders, consequences, and moral principles. Application: Analyzes Altman's decision-making processes and ethical considerations. Reference: Rest, J. R. (1986). "Moral Development: Advances in Research and Theory." Praeger.

#### Corporate Governance Principles Description: Guidelines and best practices for managing and governing a corporation, focusing on transparency, accountability, and stakeholder interests. Application: Assesses Altman's alignment with good governance practices and his impact on OpenAI's organizational stability. Reference: Cadbury, A. (1992). "Report of the Committee on the Financial Aspects of Corporate Governance." Gee and Co. Ltd.

#### Cybersecurity Risk Assessment Frameworks Description: Methodologies for identifying, analyzing, and mitigating cybersecurity risks, particularly in high-tech environments. Application: Evaluates the potential cybersecurity risks associated with Altman's actions and OpenAI's technologies. Reference: National Institute of Standards and Technology (NIST). (2018). "Framework for Improving Critical Infrastructure Cybersecurity."

#### Behavioral Economics Description: Studies the effects of psychological, cognitive, and emotional factors on economic decisions. Application: Understands how Altman's personal motivations and cognitive biases might influence his strategic decisions. Reference: Kahneman, D., & Tversky, A. (1979). "Prospect Theory: An Analysis of Decision under Risk." Econometrica, 47(2), 263-291.

### Supporting Expertise

* *Behavioral Psychology*: Expertise in understanding human behavior, personality traits, and decision-making processes. * *Corporate Law and Governance*: Knowledge of corporate structures, governance frameworks, and ethical standards. * *Cybersecurity*: Understanding cybersecurity threats and risk management strategies. * *Ethics and Compliance*: Proficiency in ethical decision-making and compliance standards.

Re: LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime

#72

Post by segmondy »

daft_pink wrote: post #39

I think the point of open cuda is to run it on non NVIDIA gpus. Once you have to buy NVIDIA gpus what’s the point. If we had true you competition I think it would be far easier to buy devices with more vram and thus we might be able to run llama 405b someday locally. Once you already bought the NVIDIA cards what’s the point

Some of us are running llama 405B locally already. All my GPUs are ancient Nvidai GPUs. IMO, the point of an open cuda is to force Nvidia to stop squeezing us. You get more performance for the buck for AMD. If I could run cuda on AMD, I would have bought new AMD gpus instead. Have enough people do that and Nvidia might take note and stop squeezing us for cash.

Re: LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime

#73

Post by jokethrowaway »

bstockton wrote: post #42
wackycat wrote: post #13

I have limited experience with CUDA but will this help solve the CUDA/CUDNN dependency version nightmare that comes with running various ML libraries like tensorflow or onnx?

My experience, over 10 years building models with libraries using CUDA under the hood, this problem has nearly gone away in the past few years. Setting up CUDA on new machines and even getting multi GPU/nodes configuration working with NCCL and pytorch DDP, for example, is pretty slick. Have you experienced this recently?

yes, especially if you are trying to run various different projects you don't control

some will need specific versions of cuda

right now I masked cuda from upgrades in my system and I'm stuck on an old version to support some projects

I also had plenty of problems with gpu-operator to deploy on k8s: that helm chart is so buggy (or maybe just not great at handling some corner cases? no clue) I ended up swapping kubernetes distribution a few times (no chance to make it work on microk8s, on k3s it almost works) and eventually ended up installing drivers + runtime locally and then just exposing through containerd config

Re: LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime

#74

Post by pplante »

candiddevmike wrote: post #62
why_only_15 wrote: post #59

Earlier quoted context omitted.

Specifically the clause is that you cannot use their consumer cards (e.g. RTX 4090) in datacenters.

That's why we run all of our ML workloads in a distributed GPU cluster located in every employee's house

The bonus is free heating for every employees household!

Re: LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime

#75

Post by btbuildem »

daft_pink wrote: post #39

I think the point of open cuda is to run it on non NVIDIA gpus. Once you have to buy NVIDIA gpus what’s the point. If we had true you competition I think it would be far easier to buy devices with more vram and thus we might be able to run llama 405b someday locally. Once you already bought the NVIDIA cards what’s the point

> Once you already bought the NVIDIA cards what’s the point

Good luck getting a multi-user GPU setup going, for example.

It super sucks when the hardware is capable, but licensing doesn't "allow" it.

Re: LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime

#76

Post by pezezin »

diggan wrote: post #30
mrbungie wrote: post #27

Earlier quoted context omitted.

Then call it Cuba Libre.

Or go even further, call it "Culo Libre", only two letters away anyways.

If nVidia can release a library called cuLitho, I don't see why not.

Re: LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime

#77

Post by amelius »

lldb wrote: post #70
amelius wrote: post #67

Earlier quoted context omitted.

By the way, can anyone explain why libcudnn takes on the order of gigabytes on my harddrive?

Primarily because it has specialized functions for various matrix sizes which are selected at runtime.

Ok, so are you saying it contains mostly straightforwardly generated code?

Re: LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime

#79

Post by candiddevmike »

marcodiego wrote: post #25
nasretdinov wrote: post #23

Such a missed opportunity to call it CUDA Libre...

Worse, some functions are prefixed libreCu which in Portuguese means "free ass hole".

I'd love to work on a codebase like that. I think codebases need more swearing to fully capture context and intent of the developer.

Re: LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime

#80

Post by cyber_kinetist »

codedokode wrote: post #34
KeplerBoy wrote: post #9

Earlier quoted context omitted.

That's a problem on a different level of the CUDA stack. Having a compiler that takes a special C++ or python dialect and compiles it to GPU suitable llvm-ir and then to a GPU binary is one thing (and there's progress on that side: triton, numba, soonish mojo), being able to launch that binary without going through the nvidia driver is another problem.

Cannot Vulcan compute be used to execute code on GPU without relying on proprietary libs? Why not?

Vulkan Compute's semantics are limited by SPIR-V and thus cannot implement all of the features CUDA provides (ex. there is no proper notion of a "pointer")

Also it's much more convenient to use plain C++ rather than a custom shading language, especially if you're writing complex numerical code or need some heavy templated abstractions to do powerful stuff. And the CUDA tooling itself is just much easier to use compared to Vulkan, with its seamless integration of host / device code.

Post reply on HN