Japan Just Built What NVIDIA Feared Most
Dipublikasikan 16 Nov 2026 · 19:02
Suka: 774Komentar: 52+24j: +69 rbRata-rata tayangan/j: 6 rb
Unggahan terbarunya
#3 dari 5 terakhir · biasanya 7 rb
Unggahannya yang kami lihat, terlama di kiri, menurut tayangan (skala log). Putus-putus: biasanya.
Performa dalam angka
Angka per 17 Nov 2026, 20.50 UTC.
Japan Just Built What NVIDIA Feared Most dari Engineered in Asia punya 174 rb tayangan, 28 jam setelah dipublikasikan pada 16 November 2026.
Sekarang mendapat 2 rb tayangan per jam, 24× dari yang biasanya didapat unggahan Engineered in Asia (7 rb).
Masuk tangga trending YouTube di 1 negara, terbaik #27 di Singapura.
Hari terbaiknya sejauh ini 17 Nov, dengan 171 rb tayangan; paling banyak ditonton sekitar 09 UTC.
| Tayangan | 174 rb | 174.488 |
|---|---|---|
| Tayangan 24 jam terakhir | +69 rb | +69.037 |
| Tayangan per jam sekarang | 2 rb | 1.842 |
| Kali biasanya channel | 24× | 24,2× |
| Biasanya channel per unggahan | 7 rb | 7.198 |
| Hari terbaik | 17 Nov · 171 rb | 171.354 |
| Jam tersibuk | 09 UTC | |
| Negara yang trending | 1 | |
| Peringkat terbaik | #27 · Singapura | |
| Suka | 774 | |
| Komentar | 52 | |
| Dipublikasikan | 16 November 2026 | |
| Durasi | 19:02 | |
| Kategori | Science & Technology | |
| Channel | Engineered in Asia | |
| Subscriber channel | 13 rb |
Tayangan
174 rb
+69 rb · 24 h
Tayangan/jam kini
2 rb
▼ 74%
× biasanya
24×
biasanya 7 rb
Belum cukup data.
Tayangan per hari
171 rb
hari terbaik, 17 Nov
Trending di
1
terbaik #27 di Singapura
Kapan ditonton
09
jam tersibuk, waktu Anda
Masuk chart di mana1
- #27Singapura17 Nov
Dari catatan YouTube
Bidang video sebagaimana dikembalikan oleh YouTube API.
- Dipublikasikan
- 16 Nov 2026, 20.45
- Durasi
- 19:02
- Kategori
- Science & Technology
- Bahasa audio
- Inggris (en)
- Bahasa judul
- Inggris (en)
- Teks
- Tidak
- Definisi
- HD
- Format
- 2D
- Lisensi
- Standard YouTube
- Konten berlisensi
- Ya
- Dapat disematkan
- Ya
- Dibuat untuk anak
- Tidak
- Dibatasi usia
- Tidak
- Promosi berbayar
- Tidak
- Statistik publik
- Ya
Topik
- Technology
Judul dan deskripsi dalam 21 bahasa
- Arab
- Bengali
- Inggris
- Hindi
- Indonesia
- Italia
- Ibrani
- Jepang
- Korea
- Malayalam
- Punjabi
- Polski
- Rusia
- Tamil
- Telugu
- Ukraina
- Jerman
- Spanyol
- Prancis
- Belanda
- Portugis
Deskripsi
Fujitsu just challenged NVIDIA with Monaca, a 144-core CPU built for AI inference that runs at 40°C on air cooling alone, slashing cooling electricity consumption by up to 80%. As power constraints threaten the global AI boom, Japan is executing an aggressive silicon strategy to break free from GPU dependence. - Fujitsu launches Monaca, using 144 onboard cores and layered memory to deliver AI inference at twice the speed of conventional CPUs. - Fugaku supercomputer shocks the industry by defeating NVIDIA-powered Summit with zero GPUs, relying entirely on Fujitsu A64FX processors. - Rapidus secures 2.35 trillion yen in government backing to build domestic 2-nanometer and 1.4-nanometer foundry lines in Hokkaido. - Preferred Networks unveils the MN-Core L1000, targeting AI answering workloads up to 10 times faster than current graphics hardware. - Jensen Huang pivots NVIDIA strategy with NVLink Fusion, opening proprietary interconnects to Fujitsu after previously dismissing rival silicon. AI compute is shifting rapidly from model training to around-the-clock inference workloads. By designing specialized silicon that bypasses graphics cards and operates without water cooling, Japan is carving out alternatives to NVIDIA's ecosystem and reshaping data center economics. Can domestic CPU designs and custom silicon truly break NVIDIA's hold on enterprise AI data centers? Drop your thoughts in the comments, like the video, and subscribe for more deep dives into the global semiconductor race. 00:00 Fujitsu Monaca AI Chip 01:24 How Nvidia Got Big 03:18 Japan Semiconductor Collapse 04:40 Fugaku Beats Nvidia 07:43 Training Versus AI Inference 09:15 Monaca Architecture Explained 11:29 Rapidus Two Nanometer Push 13:26 Preferred Networks AI Rivals 15:28 Nvidia Opens NVLink Fusion 17:45 Japan AI Independence StrategyTampilkan semua
Fujitsu just challenged NVIDIA with Monaca, a 144-core CPU built for AI inference that runs at 40°C on air cooling alone, slashing cooling electricity consumption by up to 80%. As power constraints threaten the global AI boom, Japan is executing an aggressive silicon strategy to break free from GPU dependence. - Fujitsu launches Monaca, using 144 onboard cores and layered memory to deliver AI inference at twice the speed of conventional CPUs. - Fugaku supercomputer shocks the industry by defeating NVIDIA-powered Summit with zero GPUs, relying entirely on Fujitsu A64FX processors. - Rapidus secures 2.35 trillion yen in government backing to build domestic 2-nanometer and 1.4-nanometer foundry lines in Hokkaido. - Preferred Networks unveils the MN-Core L1000, targeting AI answering workloads up to 10 times faster than current graphics hardware. - Jensen Huang pivots NVIDIA strategy with NVLink Fusion, opening proprietary interconnects to Fujitsu after previously dismissing rival silicon. AI compute is shifting rapidly from model training to around-the-clock inference workloads. By designing specialized silicon that bypasses graphics cards and operates without water cooling, Japan is carving out alternatives to NVIDIA's ecosystem and reshaping data center economics. Can domestic CPU designs and custom silicon truly break NVIDIA's hold on enterprise AI data centers? Drop your thoughts in the comments, like the video, and subscribe for more deep dives into the global semiconductor race. 00:00 Fujitsu Monaca AI Chip 01:24 How Nvidia Got Big 03:18 Japan Semiconductor Collapse 04:40 Fugaku Beats Nvidia 07:43 Training Versus AI Inference 09:15 Monaca Architecture Explained 11:29 Rapidus Two Nanometer Push 13:26 Preferred Networks AI Rivals 15:28 Nvidia Opens NVLink Fusion 17:45 Japan AI Independence Strategy
◆ Dihitung oleh viralshooter · semua angka lainnya dari YouTube API