Search results

Do you know the Supermicro cable PN?

I’m on travel so can’t see it right now but I couldn’t make this work and I’m sure I’m doing something incredibly stupid!

I’m sure I’m being stupid here but SMC chassis have a different connector for the front panel stuff than Asrock has pin headers on MB. I can’t just connect some of the pins since the SMC ribbon cable is narrower (16pins?) the the MB header (20pins?). What am I missing?

Given I just spent 2 weeks with SMC support trying to fix my issue, and I buy from them because of their Nvidia relationship, this is disappointing! Thx!

I found some outrageously expensive thing that I didn’t buy. What’s happening in July? For now I bought new CX7 variants with cages compatible with my switch. Supermicro really screwed me over here.

I’ve got one too. Same question except I want to stuff this full of double width GPUs. Can’t decide on a motherboard. The standoffs will fit anything since you can move them around. And I’m not too concerned about the built in connectors lining up—they won’t. Can deal with that by adding an...

I built one of these. It will hold 8 data center GPUs, dual slot, but it won’t 4090.

Thank you!

I only want to use NVMe anyway so this is helpful. This? https://a.co/d/iBKvuHs

I’m trying to adapt an Asrock Altra motherboard to an SMC 2U chassis that has the BPN backplane. The MB has Oculink and 4 Slimline x8 connectors for NVMe PCIe cards. But nothing else that seems relevant. I thought there might be an adapter for the BPN’s Mini-SAS outputs to something on the MB...

Can’t be done. Had to downshift to Cx6.

Anyone found a suitable 1U chassis?

Has anyone found a box that this MB fits nicely? 1U, 2U, etc work for me as I’m a bit desperate.

Where do you all buy dual slot 4090s? I can’t find them anywhere.

Or you can implement simple job scheduling and finetune on customer data during off-peak. Keep GPUs earning $ and enjoy a cost-of-money capital base instead of paying hyperscalers.

We (stardog.ai) use both but L40S for production and 6000 Ada for developers. But for our workload GH200 is 5x better than L40s for $ per 1k tokens per second. That’s on unoptimized inference stack so I expect that to creep up to 6 or even 7x better. YMMV.

Got new GH200 with CX7 in a rack with an older switch temporarily. But I can’t figure out how to cable this… I can’t find an OSFP to 4-port QSFP28 breakout except one for an absurd $2500. I need 1.5m and passive/DAC etc. Help?

I really like Asrock. And I don’t see many 22110s around so I can live with that. How is BIOS?

Not sure how I forgot about Paul. Perfect. What’s the sanest way to create more PWM headers?

I’ve been building GPU workstations for my AI developers and using Tailscale to form an R&D mesh. Good times! Most of these systems have BMC headers and I realized having IPMI/BMC/Redfish could be very useful. I can’t to save my life find an expansion card or chip to add to these systems...