Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
phonon
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
21 ms
·
271.
▲
by
phonon
2y ago
The Asus screen is very close (only significant difference is brightness), AND a touchscreen! https://www.youtube.com/watch?v=JnJw54oyfLE 14.0-inch, 3K (2880 x 1800) OLED 16:10 aspect ratio, 0.2ms response time, 120Hz refre
272.
▲
by
phonon
2y ago
The cheapest Apple with an M3 Pro is https://www.apple.com/shop/buy-mac/macbook-pro/14-inch-space... which is $1999 (with 18 GB RAM and 512 GB SSD). Full pricing on Lunar Lake is not available yet, but for ex
273.
▲
by
phonon
2y ago
It's trading significantly below book? $120 B vs $82 B. Downside is that most of their assets are their fabs...
274.
▲
by
phonon
2y ago
Possibly Battlemage running at 100% will use more power than M3's GPU running at 100%...it will take some detailed testing to track that+ Lunar Lake can be set at different TDP's (plus performance settings, on battery vs connected
275.
▲
by
phonon
2y ago
Intel Arc runs FP16 at 2:1 compared to FP32, and Battlemage on Lunar Lake is the same, and XMX FP16 is actually at 8:1. I don't think M3's GPU has a better ratio. Of course there are many other aspects, but given it's Intel&#
276.
▲
by
phonon
2y ago
Gaming is essentially all done with FP32...so it's by far the best figure of merit (excluding issues with say RDNA3 dual issue which is rarely achieved in practice). You will see gaming benchmarks come out soon, and Lunar Lake will be
277.
▲
by
phonon
2y ago
You can start here... https://www.cpu-monkey.com/en/cpu_benchmark-bench_11 M3 is around 3.5 TeraFLOPS, Lunar Lake is 5.2-6.5 TeraFLOPS. I'm sure more detailed benchmarks will be coming up soon, but realistically,
278.
▲
by
phonon
2y ago
Snarky and incorrect. This should compare very favorably (performance and efficiency) with Apple's M3 chip. Slightly worse CPU ( https://browser.geekbench.com/v6/cpu/compare/7483669?baselin... ), much be
279.
▲
by
phonon
2y ago
Bingo! Watch and be afraid.... https://youtu.be/s-vJcOfrvi0
280.
▲
by
phonon
2y ago
Zepbound is $399-$650 per month, ordered direct, if insurance does not cover it. Not great, but at least a start..... https://www.help.senate.gov/chair/newsroom/press/news-sander...
281.
▲
by
phonon
2y ago
Arguably, TSMC being Apple's preferred fabricator is why it has a current market cap 10x larger than Intel. Apple probably ends up getting chips only slightly above cost from TSMC. (It pays $18 B a year, which is pretty low for the CPU
282.
▲
by
phonon
2y ago
You only need to solve this once though?
283.
▲
by
phonon
2y ago
https://www.bastillepost.com/global/article/4076639-japans-r... "Japan's rice stockpile has dropped to its lowest level in this century due to a prolonged heatwave in 2023 and rising domestic demand, cau
284.
▲
by
phonon
2y ago
With something like https://github.com/cberner/raptorq you can do several gbits/s over high latency/lossy UDP.
285.
▲
by
phonon
2y ago
https://pgroonga.github.io/ja/
286.
▲
by
phonon
2y ago
This has already been debunked. "As it turns out, however, EUV monopolist ASML and its optics suppliers Zeiss have already considered similar setups. “Zeiss has several designs available with similar/better performance than the Sh
287.
▲
by
phonon
2y ago
You're double counting, since Geekbench 6 has three subtests that use AVX-512. ("Object Detection", "Background Blur" and "Photo Library") https://www.geekbench.com/doc/geekbench6-benc
288.
▲
by
phonon
2y ago
https://crosleylaw.com/blog/texas-personal-injury-law-change... .
289.
▲
by
phonon
2y ago
I'm not sure I understand your point. Yes, likely if NeXT had been doing better financially Apple couldn't have afforded it. You can check NeXT's aborted S-1 here.[1] They had an accumulated deficit of $273 million as of a fe
290.
▲
Why Is Japan So Weak in Software? [video]
(youtube.com)
18 points
by
phonon
2y ago
|
13 comments
291.
▲
by
phonon
2y ago
NeXt was not doing well financially, and neither was Pixar until Toy Story came out. He also did not run Pixar in any meaningful way.
292.
▲
by
phonon
2y ago
Wait until Lunar Lake comes out.
293.
▲
by
phonon
2y ago
Yes, no, yes.
294.
▲
by
phonon
2y ago
That's the point. They have more employees than the #1 silicon fabricator, the number #1 GPU designer, and the #2 CPU and GPU designer, COMBINED.
295.
▲
by
phonon
2y ago
You need an integer number of votes, so you would expect it cannot be equal to an exact fraction. But being within a single vote bound is incredibly suspicious.
296.
▲
by
phonon
2y ago
Make an ethics complaint to the state board.
297.
▲
by
phonon
2y ago
https://simpleflying.com/sukohi-superjet-100-crash-russia/
298.
▲
by
phonon
2y ago
Russia's GDP is a bit more than half of California's. Civil aviation there is a few hundred planes (about the size of Southwest--in total), and is heavily government subsidized. https://www.reuters.com/business
299.
▲
by
phonon
2y ago
What are you correcting? 59% from M1->M4 = 1.59^⅓ = 16.7% increase per generation No one said "years" either...
300.
▲
by
phonon
2y ago
Hyperthreading is turned off on m7a/c7a/r7a. "One of the major differences between M7a instances and the previous generations of instances, such as M6a instances, is their vCPU to physical processor core mapping. Every vCPU o
More ›