ycui1986

8 ms·

1.
by ycui1986 2mo ago
exactly. we need branch predictor for expert weight.
2.
by ycui1986 2mo ago
There are a lot of SSD streaming engines these days. But few to actually try some hard features. There is one that could really improve the speed. Given almost all major models come with MTP head for speculative decoding. The same MTP head
3.
by ycui1986 2mo ago
i always thought Ryzen AI Halo, together with DGS Spark, has mismatched compute capacity with memory size. Given 128GB VRAM, people would want to run large models, but the GPU compute is constraint in these types of use case. If the box run
4.
by ycui1986 5mo ago
So, dual RTX PRO 6000
5.
by ycui1986 5mo ago
I really like the pro version. The pelican is so cute.
6.
by ycui1986 5mo ago
32GB RAM on mac also need to host OS, software, and other stuff. There may not even be 24GB VRAM left for the model.
7.
by ycui1986 5mo ago
just because a bunch of rockets went up without blowing up, does not mean they are profitable. it cost money to shot rocket, and it is very expensive, reusable or not. most launches are internal launch without external paying customers.
8.
by ycui1986 5mo ago
another 60 billion to save a failed AI endeavor.
9.
by ycui1986 5mo ago
outputting docx files does not have much to do with model capability. it is about whether tool calling has be configured .
10.
by ycui1986 5mo ago
There are also many Chines AI-target GPU/NPU producers. You can get a hold of some boards on taobao.com. They are usable in some way. No, nVidia and AMD are not the only ones benefiting.
11.
by ycui1986 5mo ago
i give it in real ubuntu, no vm, no docker. so long I don't ask it to organize files, it will behave. it has not screw me so far.
12.
by ycui1986 5mo ago
qwen3.5 and qwen3.6 are both good at tool calling.
13.
by ycui1986 5mo ago
For many LLM load, it seems ROCm is slower than vulkan. What’s the point?
14.
by ycui1986 5mo ago
he won't. if anything, openai is falling behind recently. the trend won't change easily. it is like the old time Netscape.
15.
by ycui1986 6mo ago
only works if the users are evenly distributed around the globe (which is likely more of less the case). if the user concentrates in on century, the token rate will be terrible.
16.
by ycui1986 6mo ago
i hope someone do a 100b 1-bit parameter model. that should fit into most 16GB graphics cards. local AI democratized.
17.
by ycui1986 6mo ago
i am guessing, without any proof, that, when one breaker fails the server lose it all, or loose two GPUs, depending on whether one connected to the cpu side failed.
18.
by ycui1986 6mo ago
9070XT provide roughly same inference performance at double the power, half the cost, as RTX PRO 4500. So this one is optimized for total BOM cost.
19.
by ycui1986 6mo ago
they could had gone with the Max-Q version RTX PRO 6000 and only require 120V circuit. 10% performance hit, but half the power. fundamentally, looks like they are shipping consumer off-the-shelf hardwares in a custom box.
20.
by ycui1986 6mo ago
for all past years, I have been told wayland is the future. but the decade long dragged out rolling out did not made much sense to me. neither did I investigate why. until today, I found out how difficult to force a 1920x1200 resolution ove
21.
by ycui1986 7mo ago
the reality is no where to get the fuel. hydrogen stations are shutting down not building up.
22.
China tests crewed spacecraft abort and rocket recovery in major lunar milestone(spacenews.com)
3 points by ycui1986 7mo ago|1 comments
23.
by ycui1986 7mo ago
China tests crewed spacecraft abort and rocket recovery in major lunar milestone
24.
by ycui1986 7mo ago
it is bizarre that a notepad app can have remote code execution. how much unnecessary function did MS add to get to this point?
25.
by ycui1986 8mo ago
If what Waymo wrote is true, this sounds more like kids fault or guardian’s.
26.
by ycui1986 8mo ago
everyone uses cellphone that transmit on the same frequency. they don't seem to cause interference. once enough lidar enters real word use. there will be regulation to make them work with each other.
27.
by ycui1986 9mo ago
used NI-GPIB on USB cost $100 on ebay. You don’t need $1000.
28.
by ycui1986 9mo ago
very impressive. better than anything on the market either NI or Keysight.
29.
by ycui1986 9mo ago
from the picture, the compressor and generator located inside the dome. the dome is filled with CO2. maintenance people have to carry oxygen tank, or they die.
30.
by ycui1986 9mo ago
i think it had something to do with CO2 can be made into supercritical state relatively easily, not for nitrogen or other common gases.

More ›