Sunghyun Park of Rebellions.ai joins theCUBE Research and NYSE Wired to examine how purpose-built memory-centric silicon enables always-on real-time artificial intelligence inference in next-generation data centers. Park outlines Rebellions' approach to optimizing inference workloads for energy and cost efficiency, describes rack-level system design and networking and highlights commercial deployments with Korean telecommunication operators that validate the architecture. They explain how production deployments with SK Telecom and Korea Telecom demonstrate high API throughput and improved operational efficiency. Park argues that the market shifts from training to inference, where performance per watt and per dollar matter most. They emphasize that memory-centric architectures, photonics-enabled networking and rack-level co-design are critical to lowering total cost of ownership. TCO reduction enables diverse deployments in cloud, edge and on-device environments. Rebellions reports production collaborations with SK Hynix and Samsung Foundry on silicon and manufacturing roadmaps and describes how rack-scale co-design and networking choices reduce power consumption and operating expense. The discussion also covers open-source software stacks such as PyTorch for inference optimization and practical factors to consider when deploying AI infrastructure at scale.
Forgot Password

Almost there!

We just sent you a verification email. Please verify your account to gain access to . If you don’t think you received an email check your spam folder.

Didn't get an email? Click here
Sign in to Cube365 Events Platform.

In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open this link to automatically sign into the site.

Register For Cube365 Events Platform

Please fill out the information below. You will recieve an email with a verification link confirming your registration. Click the link to automatically sign into the site.

* Indicates a required field
Already Registered? Get a verification code

You’re almost there!

Didn't get an email? Click here
Profile
person_add
CONFERENCE PASS
  • General
Delete my registration
Delete my CUBE365 account
I want my badge and interests to be visible to all attendees.
Checking this box will display your presense on the attendees list, view your profile and allow other attendees to contact you via 1-1 chat. Read the Privacy Policy. At any time, you can choose to disable this preference.
Select your Interests!
add
Upload your photo
 Uploading..
OR
Connect via Twitter
Connect via Linkedin
EDIT PASSWORD

Cancel

Incorrect password


You are already logged into TheCUBE Network as

theCUBE.net
home Programs chevron_right AI Factories
  • Sign In

Share

Share clip

The clip will play by default when shared. The entire video and transcript is also accessible via the link.
Or share link
https://www.thecube.net/events/share/undefined content_copy
Manage Access
Are you sure you want to remove access rights for this user?
Details
Manage Access
email address
Community Invitation
  • More from AI Factories

Share clip

The clip will play by default when shared. The entire video and transcript is also accessible via the link.
Or share link
https://www.thecube.net/events/share/undefined content_copy
35:04
01:01
16:34
25:11
19:14
26:39
20:52
27:15
22:51
23:36
28:37
20:03
25:21
26:07
20:01
16:11
18:42
17:54
16:37
17:37
14:42
21:59
24:27
15:05
25:34
20:26
18:26
34:16
27:16
20:27
14:59
20:29
20:40
23:18
24:48
14:40
18:20
26:23
23:31
20:13
24:31
10:13
25:52
19:35
26:01
21:28
29:05
20:40
23:04
28:08
17:06
23:40
26:05
25:17
24:44
25:05
20:09
16:15
23:14
25:16
48:28
19:02
21:54
20:58
14:55
11:44
19:23
19:27
17:44
19:56
16:28
14:29
21:21
14:59
23:15
20:43
18:44
20:27
48:57
34:03
40:19
11:57
19:00
33:39
16:25
12:17
19:30
12:25
17:12
28:27
42:47
23:40
19:23
23:40
13:55
16:10
20:47
11:24
33:01
18:52
15:33
17:39
17:07
33:44
24:08
21:42
14:33
17:38
26:22
17:20
23:05
27:06
20:29
19:51
20:59
14:52
22:51
11:43
21:37
15:50
23:55
28:13
10:03
18:01
25:15
12:00
17:39
18:31
18:05
18:44
22:05
25:41
24:18
18:33
20:27
24:13
21:49
21:18
11:23
21:59
  • Info
Sunghyun Park, Rebellions
June 23, 2026
Sunghyun Park
Co-founder & CEO Rebellions AI
In this interview from theCUBE + NYSE Wired: AI Factories - Data Centers of the Future, Sunghyun Park, co-founder and chief executive officer of Rebellions, joins theCUBE's John Furrier to discuss how purpose-built silicon is powering the shift from AI training to large-scale inference. Park explains that inference — not training — is the defining force reshaping AI infrastructure, driving demand for always-on AI running at billions of tokens per day. He details how Rebellions' memory-centric architecture, 100% optimized for inference workloads, delivers subs... Read more
close
Watch video in context
bolt Powered by CUBE AI