Microsoft bloqueia truque no Windows 11 que acelerava o desempenho de discos NVMe

A Microsoft decidiu bloquear uma alteração no registo do sistema operativo que permitia aos utilizadores ativar o suporte nativo para unidades de armazenamento

TugaTech

@deutrino personally I think that #Monero is by far the least worst #Cryptocurrency and anyone who thinks '#ProofOfStake is less problematic than #ProofOfWork' never made ends meet working for a company that deals with #Ethereum, #Solana and that whole shitshow.

  • Like this shit is way more #WastefulCmputing and it's not just creating a shitload more CPU load, but literally burns through #NVMe - #SSD|s, uses a shitload of #RAM and uses a so much #Bandwith to the point that a single node can easily saturate a 10Gbit/s NIC.

Besides, any money transaction that doesn't go through #AMEX, #VISA, #MasterCard, #PayPal, … is not generating revenue for genocidal #Billionaires & #TechBros like #PeterThiel that want to literally kill everyone but their own Billionaire Caste since they believe they can just upload their minds like they have the Didact'a Composer…

Tech Billionaires Want Us Dead

YouTube

Konečně jsem dokončil dlouho plánovaný příspěvek na mém blogu:
UGREEN 20 Gbps NVMe/SATA M.2 case + WD_BLACK SN7100
https://jakubcabal.cz/blog/20260322-ugreen-nvme-m2-case-wd-sn7100/

#blog #usb #nvme #ugreen #wd_black

Never buy #ADATA / #XPG products. They suffer from #componentlottery issues, meaning they will switch components mid production if something cheaper or more convenient comes up. When your #NVMe #SSD eventually breaks down, the replacement under #warranty process is a fucking nightmare.

To compare, my NAS is using #Toshiba HDDs, and when one of them died, I quickly filed a replacement request with Toshiba, giving them the serial number of the disk and some SMART diagnostics. A shipping label was issued, and the thing was sent. Replacement arrived some weeks later.

ADATA on the other hand requires a tedious registration process, actual proof of purchase and payment for shipping to #Taiwan. This process ends abruptly, since I can't seem to find any proof of purchase anywhere. In short, the remaining 3 years warranty is fucking useless, because I forgot where I bought it.

Fuck you, ADATA.

Never buying your shit again.

PS: Backup, people. It's a thing.

#siliconlottery

DB crawling despite CPU/RAM upgrades? 🐢

Stop blaming code. Big Cloud is throttling your SSDs with the "Provisioned IOPS" tax. ☁️🛑

Escape the trap:
✅ NVMe (64k queues) vs SATA (32)
✅ Direct PCIe access
✅ 1M+ raw IOPS

Scale on iRexta Bare Metal. ⚡

🔗 https://www.irexta.com/blogs/iops-vs-throughput-storage-performance/

#NVMe #DevOps #Database #iRexta

Nvidia greenboost: transparently extend GPU VRAM using system RAM/NVMe

https://gitlab.com/IsolatedOctopi/nvidia_greenboost

#HackerNews #Nvidia #greenboost #GPU #VRAM #NVMe #systemRAM #technology

Ferran Duarri / nvidia_greenboost · GitLab

GitLab.com

GitLab

My otherwise rock solid #Lenovo X1 Gen 12 has been my trusty companion for almost six months now, but lately it started getting a little quirky.

Turns out the #ADATA 2TB #NVMe stick I immediately swapped in, is now complaining about unreadable LBA sectors. Zero percent surplus sectors used. Still keeps failing at the same couple of sectors.

I'm off to #kubecon in about a week, and I'm not dragging a dying laptop with me. Being in reasonable budget mode, I replaced the faulty stick with an older and probably a lot more trustworthy Samsung Evo stick. Half the capacity, but I doubt I'll miss it.

Debian Testing is happily installing away, and we should be back to normal shortly. Most code and data is shared with git and configuration is mostly scripted, so it's all good.

Scored this #NVMe #SSD for € 0,26 per GB ... which I think is a lucky deal in these dire times... 😎

12GBのVRAMで31GBのLLMが動く:コンシューマー向けGPUのVRAM限界をシステムRAMで補うオープンソース技術「GreenBoost」が登場

近年、オープンソースの大規模言語モデル(LLM)の性能向上は著しく、数百億から数千億パラメータクラスのモデルが次々と公開されている。しかし、これらの高度な推論能力をローカル環境で活用しようとする開発者や研究者の前に立ちはだかるのが、物理的なGPUのビデオメモリ(VRAM)容量の壁である。例えば、約31.8GBのメモリ領域を要求する「glm-4.7-flash:q8_0」のようなモデルをコンシューマー向けのハードウェア単体で稼働させることは、これまで極めて困難を極めた。 従来の解決策は主に2つのアプローチに依存していた。1つは、あふれたニューラルネットワークのレイヤーをCPU側のシステムメモリにオフロードする手法である。しかし、CPUのシステムメモリはCUDAコヒーレンスを持たないため、GPUとCPU間で複雑かつ大量のデータやり取りが発生し、深刻なボトルネックとなる。事実、トークン生成速度はVRAM単体で実行する場合と比較して5倍から10倍の速度低下を引き起こし、実用的なインタラクションを損なう。もう1つの手段は、モデルの量子化レベルを極端に下げることであるが、これは推論能力や論理的思考能力の著しい劣化を伴うトレードオフが存在する。もちろん、48GB以上の広大なVRAMを備えたエンタープライズ向けのGPUを導入するという選択肢もあるが、これはワークステーション一式を上回るコストを要求し、個人の開発者や資金力に乏しいスタートアップにとっては現実的とは言い難い。 こうした構造的なジレンマに対し、独立系オープンソース開発者であるFerran […]

https://xenospectrum.com/nvidia-greenboost-vram-expansion/

Want a better idea? Rather than replace flash modules on your #MacbookNeo, get a NAS:-) The Beelink ME Mini can host 6 m.2 #nvme drives and #TrueNas community edition can make that appear like a drive or serve your files in many, many ways. The n95 version of the ME Mini is, just fine. Very fine, even.

Nevertheless, nothing but respect here for modders! #Apple

https://wccftech.com/macbook-neo-1tb-ssd-mod/