Demand outstrips supply by such a wide margin it doesn’t matter one bit how they compare.
The cheapest 8x A100 (80GB) on the list is LambdaLabs @ $12/hour on demand, and I’ve only once seen any capacity become available in three months of using it. AWS last I checked was $40/hr on demand or $25/hr with 1 year reserve, which costs more than a whole 8xA100 hyperplane from Lambda.
I was very hopeful when I saw "AMD Support" listed on a few of those providers, but that appears to only refer only to AMD CPUs. It is, unfortunately, very difficult to find public cloud providers for AMD GPUs.
There are several tier-two clouds that offer GPUs but I think they generally fall prey to the many of the same issues you'll find with AWS. There is a new generation of accelerator native clouds e.g. Paperspace (https://paperspace.com) that cater specifically to HPC, AI, etc. workloads. The main differentiators are:
- much larger GPU catalog
- support for new accelerators e.g. Graphcore IPUs
- different pricing structure that address problematic areas for HPC such as egress
However, one of the most important differences is the lack of unrelated web services related components that pose a major distraction/headache to users that don't have a DevOps background (which AWS obviously caters to). AWS can be incredibly complicated. Simple tasks are encumbered by a whole host of unrelated options/capabilities and the learning curve is very steep. A platform that is specifically designed to serve the scientific computing audience can be much more streamlined and user-friendly for this audience.
There's a clear distinction between these two modes of transit which is suspiciously absent in this comment: A car or truck is a quickly moving multi-ton metal box with a stopping distance of ~40 feet at only 20mph. As such, a vehicle is an immediate threat to anything around it. In contrast, bicycles may pose an indirect threat to others e.g. causing an accident (a threat posed by vehicles as well) but their immediate threat i.e. loss of life due to collision is essentially zero.
It's also worth noting that drivers tend to be less alert than bikers. Distracted drivers (e.g. drivers on their phone) are a major cause of vehicles running stops signs/lights. When bikers run stop signs/lights, they are generally very aware since their lives depend on them being aware -- they are acting with intention.
In any case, for these reasons and others, there are states that actually observe what is called an Idaho Stop for bikers which allows them to roll through stops signs and lights if it is safe to do so.
> This article compares the wages not of all CEOs, but of 300 of the best paid ones.
That is not true. The report analyzes “publicly held companies with the lowest median worker pay.” They describe the sample probably a dozen or so times in the report.
> tiny fraction of one side to millions
You are making this out to be a random sample of irrelevant companies. These are publicly traded companies of which there are a few thousand in total. The list includes companies such as Amazon, Target, Best Buy, Lowe’s, Estee Lauder, etc.
Me: "This article compares the wages not of all CEOs, but of 300 of the best paid ones."
You: "That is not true."
The article: "Report on 300 top US companies...."
Tell me again this is not 300 of the best paid CEOs? After all, the fact is CEO average pay is only 212k. BLS is a much better source than the Guardian for salary data, after all.
>You are making this out to be a random sample of irrelevant companies.
It is most certainly not a random sample of companies, otherwise the result would mimic the actual data from BLS. This is specially selected data to get this result.
It is a cherry picked sample of the very top end of the CEO salaries (300 of them out of over 200,000 of them that exist) compared to workers, but not all workers, only those that are among the lowest paid, again, to reach a shocking headline that creates outrage.
Here's [1] the actual report. Note in the report they compare part time worker wages to fulltime CEOs - not an hourly rate or something a tiny bit more defensible - for example, they compare an annual wage of $224 (yes, you read that correct) for a part time worker at NuSkin to the CEO wage of $4m. They repeat this type of nonsense throughout the report. The dishonesty in the report to me is astounding - and it explains how they get the numbers to drive their headlines.
Comparing a tiny fraction of top values from one dataset to all of another dataset is not exactly a valid statistical technique. Doing everything you can at each step to widen the result to get what you want is beyond simple stats misunderstanding - it's nutjob agenda driven lying.
If someone did this to claim global warming was not real, or any of a zillion other false claims, people would (and should) cry foul.
If I did this, picking the top 300 employee wages in the country (all well into the millions) and compared to all CEOs (avg pay 200k), obtaining the claim that workers make many multiples of CEO wages, you'd rightfully cry foul.
So why the pitchforks and intellectual dishonesty when the data is manipulated this way? Because it gets a result you want to see?
Your balance sheet is either cash, which can evaporate overnight due to rampant inflation which many economists are predicting, or you can convert some or all of that to BTC. As a company, you need to manage your cash. Companies invest their balance sheet in debt, money markets, etc. That doesn’t imply that you are somehow a fund. You have no choice but to manage your cash. I recommend just reading Michael Saylor’s posts or listen to this podcast which provides a good summary of their thinking: https://open.spotify.com/episode/1VwUjvMNeoOeyiQwP3im6G?si=z...
Bitcoin has had a 50% drop two times in the last three years. No matter how bullish you are on its long-term prospects, it can't possibly be seen as a stable store of value in the short term.
At this point they are effectively staked their entire success to BTC. They aren’t just putting cash into BTC they are taking on additional debt to buy BTC.
If you had a strong core business anyway you wouldn’t do this, it’s just a sign that their core business is failing and that pivoting to a Bitcoin investment company is a Hail Mary.
The question is why a company doing < $500mm in annual revenue is holding $4B in cash. That's really the only question. It doesn't matter whether it's crypto or real estate or USD. The question is why it's holding so many excess assets at all.
100% agreed that Kubernetes is overkill for many if not most deployments. We constantly see small startups prematurely adopting Kubernetes which is a costly investment in terms of building up internal knowledge and maintaining the cluster. Gradient (https://gradient.paperspace.com) is push to deploy service built on Kubernetes but you don't need to know anything about Kubernetes to use it. We feel like is the right way to leverage the power of Kubernetes unless you're at Netflix or Lyft scale. In Gradient, you just provide a model, you select an instance type (several affordable GPU options offered), and a docker container. Everything else e.g. autoscaling, auth, rolling updates, etc. is handled automatically. Kubernetes does an amazing job providing the backend for these operations but data scientists and even devops teams at startups should not be wasting time rolling their own Kubernetes cluster and installing/maintaining an inferencing service on top.
They are just two different solutions that have pros and cons just like any two solutions :) A few that jump out:
- Setup time: Setting up GCP, setting up a certificate, adding a static IP, etc. is not seamless/adds friction
- Autoscaling and rolling updates (no downtime)
- Team management and collaborative environment with usage tracking, permissions, etc.
- Optional integration with a pipelining service for training, tuning, deploying models in a single tool
And a point of clarification: Practically speaking, neither tool is free. Both require a cloud instance so they will cost roughly the same for the end user (Gradient also supports preemptible instances).
It’s poorly worded but I believe the OP was using an example where both the right and the left abuse a system for control and not necessarily the greater good. While it’s true that both the left and the right both abuse gerrymandering, I don’t believe that the latter part of the assertion that they do so purely in the pursuit of power/control is accurate or a fair representation of everyone’s motives. There are people involved in politics that truly believe in what their party stands and moreover, that the opposing party will induce some sort of harm (regardless if they’re well intentioned or not) if their policies are implemented. There is no law against Gerrymandering and this objectionable tactic will invariably be used by the opposing party. By not Gerrymandering, you are giving the opposing party a distinct advantage and the opportunity to enact policies that you are fighting against. The cynic would say that everyone involved cares only about power and while that is almost certainly true for certain individuals, it’s not true for many. It concerns me how some bad apples, opportunists, and some folks with an agenda in politics, the media, etc. have turned so many into cynics. There are many good people in politics and the media — people who care deeply about their work and their impact on society. We should find ways to discuss specifics and not in generalities.
It will never cease to amuse me that "bad apples" is thrown around like that, downplaying the problem. And then people nod and are like, "yeah, that vaguely reminds me of a folksy saying, it must be credible."
The saying is "one bad apple spoils the barrel", and is stressing how much of a problem it is for there to be individual deviance, and how we must be super vigilant for it and throw it the heck out. If we're to believe that, and we're to believe there have been "a few bad apples" for some time, we should expect that the whole barrel is spoiled and would be quite right to be cynics.
I don't necessarily agree, and I certainly don't think that folk wisdom is the best way of making decisions... but it's probably better than the uncritical opposite!
What exactly do you expect to learn from debating anti-semites, self-proclaimed Nazis, or people that are shouting “hang Mike Pence”? https://youtu.be/Fag0aC_M0_U
Today’s news:
Joe Biden condemned the rioters who stormed the Capitol as “a bunch of thugs,” “domestic terrorists” and “white supremacists.” The president-elect specifically called out the rioters who wore shirts saying “6MWE.” “6MWE” is an anti-Semitic phrase that stands for “Six million wasn’t enough,” referring to the six million Jewish people who were murdered during the Holocaust.
If you feel like we are where we are because we’re not listening to these people than you can count me out. I have absolutely nothing to learn from a white supremacist or someone inciting violence. And likewise, they have absolutely nothing to contribute to political discourse.
