Add Nvidia 4.1 inference code support

[anandhu-eng]

---

Migrated from mlcommons/mlperf-automations_archived#23
Originally created by @arjunsuresh on Fri, 13 Sep 2024 10:34:29 GMT

---

null

[anandhu-eng]

---

Migrated from mlcommons/mlperf-automations_archived#23 (comment)
Originally created by @Oseltamivir on Sun, 27 Oct 2024 18:05:20 GMT

---

Ah. I will try it on my 3090 rig. If that doesn't help, i'll revisit a working implementation on a H100 server.

I will update that issue if i have any breakthrough

[anandhu-eng]

---

Migrated from mlcommons/mlperf-automations_archived#23 (comment)
Originally created by @Oseltamivir on Sun, 27 Oct 2024 17:40:46 GMT

---

Hi, I am interested in the progress on this and I'm curious if i can help implement this.

[anandhu-eng]

---

Migrated from mlcommons/mlperf-automations_archived#23 (comment)
Originally created by @arjunsuresh on Sun, 27 Oct 2024 18:02:01 GMT

---

Thank you @Oseltamivir for your interest. Currently we are blocked by this issue which is preventing us from checking Nvidia code on a custom system.