DeepSeek just open-sourced its inference engine

Re:DeepSeek just open-sourced its inference engine by xrisk on Tuesday October 20, 2026 @06:32AM (#50164052)

For comparison, rental prices seem to be ~$2.75/ per H-100-hour [1]. Meanwhile, Deepseek says "38 USD per hour per card”. They haven’t specified which card as far as I can tell.

[1] https://www.ft.com/content/a1815bc5-b5d5-47ff-bbbc-e1a689929...

[ Parent | on HN ]

Re:DeepSeek just open-sourced its inference engine by d5lt5 on Tuesday October 20, 2026 @06:35AM (#50164091)

It's specified in the plain text:

Supported hardware:

NVIDIA SM100 GPUs

Huawei Ascend 950 NPUs

[ Parent | on HN ]

Re:DeepSeek just open-sourced its inference engine by xrisk on Tuesday October 20, 2026 @06:42AM (#50164163)

that doesnt say what cards the reference number used.

[ Parent | on HN ]

Re:DeepSeek just open-sourced its inference engine by d5lt5 on Tuesday October 20, 2026 @07:02AM (#50164393)

Are you looking for serial numbers or something? GPU family is enough to get the cost estimate.

[ Parent | on HN ]

Re:DeepSeek just open-sourced its inference engine by azurezyq on Tuesday October 20, 2026 @01:20PM (#50169634)

SM100 -> Blackwell

[ Parent | on HN ]

Re:DeepSeek just open-sourced its inference engine by zurfer on Tuesday October 20, 2026 @06:41AM (#50164149)

It's harder than ever to get h100s and then you have to usually commit to a long time period upfront, but yeah this release should close some of the gap.

[ Parent | on HN ]

Re:DeepSeek just open-sourced its inference engine by GreenWatermelon on Tuesday October 20, 2026 @06:50AM (#50164257)

I also assume they will just slot in whatever cards they can get their hands on, so this figure could be an average.

[ Parent | on HN ]

Re:DeepSeek just open-sourced its inference engine by Schlagbohrer on Tuesday October 20, 2026 @09:06AM (#50165808)

I would love to read the details about how datacenters do load managing across such diverse hardware, if in fact they really are plugging in any computer hardware they can get on the market. Like how do they balance and serve an even customer experience across a wide range of hardware... is it by clustering similar hardware together?

[ Parent | on HN ]