Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
exged
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
6 ms
·
31.
▲
by
exged
6y ago
Once you take SIMD into account, lookup tables frequently lose. Vector gathers are not cheap at all.
32.
▲
by
exged
6y ago
What you want for asymmetric cryptography is " always hard to solve, easy to verify". Only a very small number of problems are believed to fall in this category, including factoring and discrete logarithm. It's far more comm
33.
▲
by
exged
6y ago
That's for model inference. For training OpenAI said they used around 3000 PetaFLOPS / days on the largest GPT-3 model. That translates to about 300 Nvidia A100 GPUs if you want training to finish in a month (any slower and your r
34.
▲
by
exged
6y ago
SVE doesn't really solve #2 and #3 by itself. You can still design a core with huge SIMD units that require a ton of power and area, resulting in downclocking and relatively poor performance on non-SIMD workloads. #1 is a big deal thou
35.
▲
by
exged
6y ago
It's pretty simple - for regular users Chrome is faster, has more features, uses less power, and has better website support. It is also aggressively marketed by the largest advertising company on the planet. It wasn't always this
36.
▲
by
exged
6y ago
ML contributes to a significant fraction of revenue at three of the world's largest companies (Amazon, Google, Facebook - largely through recommendations and ad ranking). It also drives numerous features that other tech companies build
37.
▲
by
exged
6y ago
Liquid hydrogen tanks are definitely a different problem, but I don't know if "orders of magnitude safer" is necessarily true. Many spacecraft have run into problems involving liquid hydrogen storage for a variety of reasons:
38.
▲
by
exged
6y ago
High-level IC roles (in the companies that have them) are still leadership roles. They are just technical leadership roles instead of people management roles. They may not have direct reports, but they are still responsible for getting big
39.
▲
by
exged
6y ago
It's desirable, but typically hardware devices have DMA-capable buses that can read / write to arbitrary physical memory. So a buggy or malicious WiFi driver may be able to use the WiFi DMA hardware to write into the GPU driver&#x