How does nvtop show the Pcie gen and speeds under load?
AI HARDWARE
-
Stable Diffusion 3.5 FP8 TensorRT Checkpoints Now Available
By
–
Stable Diffusion 3.5 FP8 checkpoints, optimized for TensorRT, are now available. ⚡ 2x AI inference speed, half the VRAM, deployable on all NVIDIA Blackwell devices. Get started ➡️ nvda.ws/4gGkfii @StabilityAI
→ View original post on X — @stabilityai, 2025-09-24 23:01 UTC
-
Gaming Hardware Inadequate for AI Server Builds
By
–
the basics: don't build an AI "Server" on gaming parts – gaming mobos = 24-48 pcie lanes, limits you to 2 gpus max before bandwidth issues – gaming cpus (i9/ryzen) are built for games, not heavy multi-gpu workloads rgb ram ≠ ecc rdimms, you want error correction when pushing
-
VINTECC Axelera Balena Edge AI Stack Partnership
By
–
VINTECC + Axelera AI + @balena_io = a powerhouse edge AI stack. VINTECC: industrial AI & digital twin systems Balena: fleet mgmt + OTA + container-native OS Axelera: high-efficiency, high-throughput inference We're always looking to connect with people using edge #AI
-
Dual Phone Strategy Gains Traction Among Tech Users
By
–
I truly considered double phoning this time and honestly might bring a Pro on trips. @davemorin was simply early on the day phone / night phone trend.
-
Sofinnova NVIDIA Alliance Scales Biotech AI Startups
By
–
💡 How do you scale 100+ biotech, medtech, and digital medicine startups—faster?
— NVIDIA AI (@NVIDIAAI) 24 septembre 2025
Simon Turner, Partner at Sofinnova, shares how their team drives:
⚡ Streamlined analytics + faster breakthroughs
🎯 Greater impact with every investment
See how Sofinnova and NVIDIA’s VC Alliance… pic.twitter.com/TFm2I2x4M1How do you scale 100+ biotech, medtech, and digital medicine startups—faster? Simon Turner, Partner at Sofinnova, shares how their team drives: Streamlined analytics + faster breakthroughs Greater impact with every investment See how Sofinnova and NVIDIA’s VC Alliance
-
Nvidia Chips as a Subscription Service Model
By
–
Actually I think that’s smart. This way you don’t have to deal with reselling and recycling old chips every few years. Nvidia chips should be offered as a subscription 😀
-
Germany’s 4,000 GPUs: Insufficient for True Digital Sovereignty
By
–
4,000 GPUs – is that supposed to secure Germany's “digital sovereignty”? While OpenAI and others are stacking millions of GPUs worldwide, Germany is celebrating a number that is barely enough to seriously train a single frontier model. Symbolic politics in high-tech costume.
-
Nvidia’s Leasing Structure Strategy for Server Collateral Financing
By
–
Hmmm. “The leasing deal could be structured to minimize risks for Nvidia. Nvidia could set up an entity that borrows money to buy the servers, using the chips as collateral. OpenAI’s lease payments could go towards paying back the loan.”
-
Data Centers and the Religious Fervor Around Super-Intelligence
By
–
Given the quasi-religious way people think about super-intelligence, I keep wondering if today's incredible data-center build out might be comparable to the construction of cargo cults.