overview for a5ehren

NVIDIA's connector story: EPS vs. 12VHPWR Connector - Unfortunately, the good doesn't always win (insights) in c/[email protected]

[–] [email protected] 1 points 9 months ago

Motherboard has way more room for connectors and power circuits than a GPU.

NVIDIA's connector story: EPS vs. 12VHPWR Connector - Unfortunately, the good doesn't always win (insights) in c/[email protected]

[–] [email protected] 1 points 9 months ago (6 children)

Just adding more and more 8-pin connectors is not a sustainable solution.

The real fix is probably to go back in time and have PCI-e 3.0 revamp the power delivery to allow more than 75W to come from the slot.

That moment when the ISP's router does better than your own. in c/[email protected]

[–] [email protected] 1 points 10 months ago

My ATT Fiber gateway is…fine. Sends WiFi 6 to (almost) the entire house and the only way I could do better is to run wires and rebuild the whole network.

Even has a 4x4 antenna array for the 5ghz band, so I would have to spend some decent money to get a better standalone router.

Please help me take advantage of this. in c/[email protected]

[–] [email protected] 1 points 10 months ago (1 children)

Yeah. Most people on this sub would kill to move into that setup, btw.

Question about cuda toolkit compatibility in c/[email protected]

[–] [email protected] 1 points 10 months ago

The website lags behind. I’m told that internally they use Wikipedia to reference CC levels.

Nvidia, Intel claim new LLM training speed records in new MLPerf 3.1 benchmark in c/[email protected]

[–] [email protected] 1 points 10 months ago

“NVIDIA was once again the only company to run all MLPerf tests. H100 GPUs demonstrated the fastest performance and the greatest scaling in each of the nine benchmarks.”