Sabitlenmiş Tweet
my cpu can now issue multiple instructions at once🙂 one step closer to being superscalar
English
Alec Petridis
638 posts
Alec Petridis
@chop0_
third-year electrical engineering student
Katılım Şubat 2021
244 Takip Edilen312 Takipçiler
Alec Petridis retweetledi
The WEF is hilarious because everything they do centres around making capitalism more sustainable and more fair to those displaced by its dynamics, but people will just fundamentally misunderstand them to make them appear purely evil.
Samantha Smith@SamanthaTaghoy
“Water, soil, and oxygen should not be infinitely accessible. They are assets that should be included in global economic balance sheets.” This is not satire. The World Economic Forum wants to monetise breathing.
English
SNCF, the french railway operator, originally was working on CAHSR but found CA so politically dysfunctional they developed a project in Morocco instead (2011). In 2018, it opened. Meanwhile CAHSR is still at least 10yrs away.
Tragic
CA High-Speed Rail 🚄💨@CaHSRA
California High-Speed Rail is creating good-paying jobs for hardworking members of the trades. 🛠️16,000+ jobs created 🏗️1,600+ daily-workers dispatched These jobs mean steady, mortgage-paying careers for thousands of Californian families. #BuildHSR
English
@akcharine It's on my GitHub: github.com/chop0/mri-stud…. If you run ./gradlew run in the mri-studio subdirectory, it should run. You can create a project, and then import the example JSON, and then you can promote some of the sequences in that to editable sequences (UX will be improved...)
English
My NMR simulation app now has DAW-like functionality for creating and visualising sequences.
English
Alec Petridis retweetledi
Engineering is actually much closer to art than it is to physics or other fundamental sciences. Physics is closer to (serious) economics than it is to engineering.
Sadly this isn't reflected in the types of people who major in art, engineering, etc. for other reasons
English
Claude mythos credits will have completely replaced the petrodollar by 2027
English
It's sad that two of my favourite projects I've worked on -- graal-bindgen and wireguard-java -- could probably be recreated with Claude in a day or two.
English
@felpix_ I think you sometimes mix up the 'is's and the 'ought-to-be's in your tweets
English
gpa matters significantly less if you went to undergrad a decade ago and have actual work/life experience
very silly behavior
Arjun Panickssery@panickssery
Interesting aspect of grade inflation I overheard is the more literal "inflation" where if you earned a strong 3.9 GPA a decade ago and want to apply to a top law school you're now not competitive at all, and indeed people are choosing not to delay law school because of inflation
English
I made a dashboard for visualising and comparing MRI excitation sequences.
English
@supr3me_lead3r @yashiro_ld Yea, I think I found a sequence that should work despite the terrible field strength and homogeneity the machine will have. I also ordered some more RF equipment
English
Red: convex hull of the magnetisation of points in the the off-slice area. Green: points in the on-slice area. If we apply a gradient field along with an x/y RF signal during refocusing, off-slice coherence improves significantly at some cost to on-slice coherence @yashiro_ld
English
An underrated consequence of AI is that academics won't need to spend 30% of their time manually measuring the pixel width of circles in ImageJ
English
@yashiro_ld It can also be used on long, multi-legged sequences to improve convergence, even with very poor field strengths and homogeneities. I am still having trouble aligning the altitude of the different isochromats, even though the azimuth (phase) is easy to align.
English
@yashiro_ld Although it's impractical to add a Golay coil on the z-axis for a linear gradient, more practical z coil geometries will be sufficient, as long as the bandwidth of the sinc refocusing pulse is set correctly.
English
@yashiro_ld is this using GRAPE, model-predictive control or something else?
English
RFパルスの最適化過程のアニメーションです。
異なるパラメータのブロッホ方程式に従う集団を同一つのRFパルスで同じポイントに再収束させます👀
Yashiro@yashiro_ld
わーい！！MRI向けに再収束RFパルスの最適化アプリできたよ！！ †解析勾配†で爆速に計算できるでヤンス🤤
日本語
@wholyv So the final output is about Vout = g_m2 * V_d * beta^2 * (1500/(1500 + beta/g_m6)) * -340 (negative sign due to the current direction across R12/R13). It'll be a bit less than this, since we assumed V_A = infinity for the BJTs and ignored Q5's degeneration.
English
@wholyv g_m * V_d is then amplifier in Q5 by beta, which is then divided across R10 (1.5k) and Q6's input resistance (beta/g_m6), and amplified in Q6 by beta again. This is then burnt across R12/R13=340 ohms. Everything else after that has a gain near 1...
English
you know what? I will just stick to software.
lyv ⌘@wholyv
i am extremely pleased to announce that I have decided to never touch software ever again in my life. I am completely shifting my focus to electronics and I am gonna love every second of it.
English