fancyfredbot

7 ms·

31.
by fancyfredbot 2mo ago
If I'm squeezing enough margin out of my customers that I can afford to sponsor a football team with some of the leftover cash then you probably don't want to deal with me.
32.
by fancyfredbot 2mo ago
It's the GNU/OSS AI bros who are upset. The other AI bros are using LLVM/Clang I expect. GCC is much less relevant than LLVM in the AI world. NVIDIA's entire CUDA compiler stack is built on LLVM, just like AMD's HIP
33.
by fancyfredbot 2mo ago
It's a pun on "efficient markets hypothesis". The efficient markets hypothesis says the market prices incorporate all available information and there's no such thing as a cheap stock, so there's no easy money to be
34.
by fancyfredbot 2mo ago
I don't entirely understand why they "didn't want to touch it"? Surely they were capable of soldering a new port. I'm sure they'd have had to charge a lot. There's unlikely to be much difference in cost be
35.
by fancyfredbot 2mo ago
Correct
36.
by fancyfredbot 2mo ago
ancient intel did 80 bit fp until your intermediate result was written from registers to memory. Since you typically have no control over when that happens, even in very low level languages like C, the 80 bit precision mostly had the effect
37.
by fancyfredbot 2mo ago
I think the term in this case is marketing. NVIDIA used to use the term supercomputer more often five or six years ago. It's now fully committed to the term AI factory. I think this is supposed to make you think of their expensive kit
38.
by fancyfredbot 2mo ago
It's mind blowing seeing these multi exaflop single rack systems. The world's first exaflop supercomputer was Frontier. It was launched only 4 years ago in 2022. It's not a fair comparison of course. FP4 in Helios barely qual
39.
by fancyfredbot 2mo ago
Pensions will have exposure, probably through private credit, and probably in the range of 5% to 10% of assets. It's a lot, and it'll hurt people. But I'm not at all sure it is enough to require a bailout. US government direc
40.
by fancyfredbot 3mo ago
Yes they are going to ramp up capacity. It would only make business sense to do nothing if all your competitors were also doing nothing, which would probably require some level of illegal collision. If everyone ramps up then in the best cas
41.
by fancyfredbot 3mo ago
I bet there's a better way than unpacking, running sequentially and repacking. Even if the algorithm is very branchy you save a pack and unpack.
42.
by fancyfredbot 3mo ago
The relatively slow depreciation of GPU value is an artifact of supply constraints. If you run fp4 inference and could choose freely between Hopper and a Rubin, the performance per watt would make the Hopper unattractive even if you paid ze
43.
by fancyfredbot 3mo ago
s/banks/private credit/
44.
by fancyfredbot 3mo ago
Things have changed since 2009. It would be private credit which fails this time, not the banks. Private credit is not supposed to be systemically important and it's not supposed to need bailing out. Maybe we'll find out how true
45.
by fancyfredbot 3mo ago
A lot of OpenAI/Anthropic compute actually belongs to the hyperscalers. They have long term contracts and will be able to charge some kind of margin for access to the compute they've secured. However I don't think this would
46.
by fancyfredbot 3mo ago
The term vibe coding actually originated from a tweet from Andrej Karpathy and so there is actually a sort of definition of the term. He said vibe coding is where you "forget the code exists": https://x.com/karpath
47.
by fancyfredbot 3mo ago
I increasingly feel that the reason vibe coding specifically prohibits reading the generated code is that it's impossible to forget the horror that lurks in these python files. The code will do what you asked for in a broad sense but w
48.
by fancyfredbot 3mo ago
Buffoon (not bafoon).
49.
by fancyfredbot 3mo ago
They are taught the difference through reinforcement learning with verifiable rewards. Pretending you've solved the task or making up a story about how you solved it won't do well in that training step.
50.
by fancyfredbot 3mo ago
Normally people refer to the compute-bound phase as "prefill". Nothing wrong with saying it's building the kv cache though, it's accurate just unusual.
51.
by fancyfredbot 3mo ago
Having Mojo support multiple platforms creates incentive to adopt Mojo and therefore write code in a language which can compile and run on Qualcomm hardware. This is good for Qualcomm. However the danger is that the language sees wide adopt
52.
by fancyfredbot 3mo ago
The top end of the range they've shown here for A2W will assume you only heat the water to a very low temperature, and have a huge underfloor radiator. If you do that, and you only heat the water to 35C, A2W can be more efficient. Most
53.
by fancyfredbot 3mo ago
Modular now joins SYCL, OpenCL, and One API on the list of cross platform languages which never really became cross platform. After so long and so much investment in AI, the best cross-platorm API we've got for high performance Kernels
54.
by fancyfredbot 3mo ago
Air to water heat pumps are typically less efficient than air to air. Both will work in cold UK temperatures. However the water is often heated to a higher temperature which can reduce efficiency.
55.
by fancyfredbot 4mo ago
Question one: How much did this cost OpenAI? Question two: Why are OpenAI spending that money taking talent from Google, who can definitely outspend them for talent, and not Anthropic, who are leading the market and are at least somewhat fi
56.
by fancyfredbot 4mo ago
I don't think my logic would lead to only one country with a launch rocket. Russia and China do not really have the option of using American rockets. Britain however does have this option and as far as I know it has never had a proble
57.
by fancyfredbot 4mo ago
The way I thought it worked was that congress would set a budget and scientists would decide how to spend it. Perhaps naively I thought these scientists would want to do science and would be unwilling to steer funds away from whichever pr
58.
by fancyfredbot 4mo ago
I understand the present US administration would want to stop funding this, and that they have the power to do so. I don't understand how that has led to the sensor network being dismantled. Surely it would have been cheaper to leave i
59.
by fancyfredbot 4mo ago
I was suggesting US models are very good, and would probably be preferable to a less capable sovereign AI . My point was that this makes building European sovereign AI unappealing. Similarly I was not suggesting Chinese models are an alte
60.
by fancyfredbot 4mo ago
Forget regulations, would it make sense for Europe to train a frontier model of it's own? Would it be sufficiently better than fine tuning a Chinese model? Would it actually be competitive with US frontier models? Would enough people p

More ›