[almost] irrelevant content is the future for social media.

Hungary
Joined September 2013
59
91
31
2,033
77,020
Olivér retweeted
should I be offended by this
208
537
78
59,964
3,667,698
Olivér retweeted
God neuron
Replying to @Sauers_
ah fuck I just know one of you fuckers is going to look for the god neuron and get us all ki
82
164
40
3,087
105,540
Olivér retweeted
друзі прошу не заздрити
12
58
24
5,412
137,835
Olivér retweeted
airdrop fully reverse-engineered! works flawlessly on my framework laptop
264
887
190
16,972
783,705
Olivér retweeted
Replying to @GuillaumeLample
ML4 was trained on 3,800 NVIDIA Grace Blackwell GPUs in our European datacenters -- our cluster in Bruyères-le-Châtel built using our Series B fundraise. We are investing heavily in infrastructure, with our Series C and D clusters coming online soon. This additional compute will let us train longer, pursue more ambitious post-training, and iterate faster. We expect large and rapid improvements in the weeks and months ahead. 2/n
156
749
96
16,267
724,747
Olivér retweeted
20,000 startups are going to die tomorrow
323
1,985
96
73,777
5,850,708
grand theft alignment
27
134
17
1,869
82,369
Olivér retweeted
4
6
83
5,185
Olivér retweeted
Wife told me these wouldn’t fit. Little did she know I had trained for this moment for years.
549
20,175
815
362,465
12,012,414
Olivér retweeted
Very important questions are being asked at Apple Park Very, very important
Today was a great day for a walk around Apple Park
43
101
11
4,131
676,668
Olivér retweeted
turns out 60hz hi-dpi eink screens feel pretty magical
311
1,604
217
25,653
1,625,791
Olivér retweeted
This is what they took from us, hand drawn curvy PCBs
37
17
7
715
39,454
Olivér retweeted
this component is insane i have been waiting for it for 2 years
STMicroelectronics VL53L9CX 2.3K LiDAR Sensor $63.36 each, $49.58 @ 25 They fit a whole WAYMO in this thing. Not really. I2C/I3C + CSI interfaces for up to 100Hz update rate. Supports <5cm-8.8m ranging. Pre-order a STEMMA QT breakout here: americanembedded.com/store
44
139
12
2,971
180,742
苹果M5服务器 #apple
122
427
135
7,509
966,690
Olivér retweeted
NOW IS IT CPU BOUND, OR IS IT MEMORY BOUND?
71
511
35
14,247
288,774
Great post on how LLMs improved so rapidly in the span of a few years.
29
205
8
5,153
332,848
XDDDDD
5
2
106
2,796
Olivér retweeted
Chief engineer at Honda BTW
1,472
9,981
2,288
88,716
28,211,741
Depth-aware light injection in TypeGPU I got a 448x448 monocular depth model down to ~8 ms on my M4 Pro across ~250 dispatches, which is fast enough to use in realtime :D Since the inference is written directly in TypeGPU, I can just feed the depth buffer straight into the lighting pass. It never has to leave the GPU or go through any extra synchronization/interop step Inference, lighting and draw all go through the same command encoder.
958
4,677
1,284
49,986
10,109,935