The Cloudcast Massive Studios

John Yue (CEO & Co-Founder @ inference.ai) discusses AI workload sizing, matching GPUs to workloads, availability of GPUs vs. costs, and more.

SHOW: 815

CLOUD NEWS OF THE WEEK - http://bit.ly/cloudcast-cnotw

NEW TO CLOUD? CHECK OUT OUR OTHER PODCAST - "CLOUDCAST BASICS"

SHOW NOTES:
Inference.ai (homepage)TechCrunch postSiliconAngle post on ChatGPUTopic 1 - Our topic for today is sizing and IaaS hosting for AI/ML. We’ve covered a lot of basics lately, today we’re going to dig deeper. There is a surprising amount of depth to AI sizing, and it isn’t just speeds and feeds of GPUs. We’d like to welcome John Yue (CEO & Co-Founder @ inference.ai) for this discussion. John, welcome to the show

Topic 2 - Let’s start with sizing, I’ve talked to a lot of customers recently with my day job, and it is amazing how deep AI/ML sizing can go. First, you have to size for training/fine-tuning differently than you would for the inference stage. Second, some just think, pick the biggest GPUs you can afford and go. How should your customers approach this? (GPU’s, software dependencies, etc.)

Topic 2a - Follow-up question what are the business side, what are the business parameters that need to be considered? (budget, cost efficiency, latency/response time, timeline, etc.)

Topic 3 - The whole process can be overwhelming and as we mentioned, some organizations may not think of everything. You recently announced a chatbot to help with this exact process, ChatGPU. Tell everyone a bit about that and how it came to be.

Topic 4 - This is almost like a match-making service, correct? Everyone wants an H100, but not everyone needs or can afford an H100.

Topic 5 - How does GPU availability play into all of this? NVIDIA is sold out for something like 2 years at this point; how is that sustainable? Does everything need to run on a “Ferrari class” NVIDIA GPU?

Topic 6 -  What’s next in the IaaS for AI/ML space? What does a next-generation data center for AI/ML look like? Will the Industry move away from GPUs to reduce dependence on NVIDIA?

FEEDBACK?
Email: show at the cloudcast dot netTwitter: @cloudcastpodInstagram: @cloudcastpodTikTok: @cloudcastpod

For some strange reason, “maintenance” has been in the news quite a bit lately. Is there ever a time when maintenance is enjoyable, or appreciated? 
SHOW: 814
SHOW TRANSCRIPT: The Cloudcast #814
SHOW VIDEO: https://youtube.com/@TheCloudcastNET 
CLOUD NEWS OF THE WEEK - http://bit.ly/cloudcast-cnotw
CHECK OUT OUR NEW PODCAST - "CLOUDCAST BASICS"


SHOW NOTES:
AWS increased the price of longer-running EKS clusters by 6xBroadcom changes VMware licensing from perpetual to subscriptionBroadcom offers security patches to perpetual license customersIncreasing the Kubernetes support window to 1 yearDiscovering the XZ backdoor (Oxide and Friends, podcast)IS MAINTENANCE EVER APPRECIATED OR ENJOYABLE?
Spent the day surrounded by maintenance activities (oil, AC, power-wash)The costs of maintenance are real and opportunityMaintenance often goes unappreciated and unseenNaming: Release Notes, Technical Debt, Chaos EngineeringTECHNICAL DEBT VS. MAINTENANCE
Should we encourage a lack of maintenance vs. innovation as a priority?Should we encourage active maintenance with lower hard costs?Is there a way to put respect on maintenance? (e.g. OSS maintainers)Do we undervalue maintenance (e.g. Backup/Recovery, DisasterRecovery, etc.)?What maintenance best practices do you use? What are the good and bad of them?FEEDBACK?
Email: show at the cloudcast dot netTwitter: @cloudcastpodInstagram: @cloudcastpodTikTok: @cloudcastpod

Kalyan Veeramachaneni (@kveeramac, CEO/Founder @DataCebo) discusses the generation and value proposition of synthetic data for GenAI.

SHOW: 813

CLOUD NEWS OF THE WEEK - http://bit.ly/cloudcast-cnotw

NEW TO CLOUD? CHECK OUT OUR OTHER PODCAST - "CLOUDCAST BASICS"

SHOW NOTES:
DataCebo (homepage)Synthetic Data Vault - SDVTechCrunch ArticleMIT News ArticleTopic 1 - Our topic for today is synthetic data. While the concept and need for synthetic data has been around for a long time, it isn’t a topic that typically comes to the forefront and something we haven’t talked about until today. Today is a bit of crossing the streams between developers and testing data and using GenAI to achieve this goal. For this, we’re joined by Kalyan, CEO and Co-Founder of DataCebo. Welcome to the show

Topic 2 - First, for those not familiar, what is synthetic data? What is the use case and need? What problem is it solving today?

Topic 2a - Hopefully, listeners out there are making the connection to the advantages of GenAI for synthetic data, but take us through your original concept at MIT and the history of Synthetic Data Vault (SDV).

Topic 3 - We recently did a show on the security and privacy of training LLMs where we covered the need to mask PII for the training of models for compliance. I can also see bias issues coming into play or maybe training data that doesn’t exist in the real world (weather models example). What are some of the use cases that you’ve seen require synthetic data sets. Are there certain industries (healthcare, financials, etc.) that benefit?

Topic 4 - You were designing this based on GenAI before GenAI was “cool”. How has the rise of LLMs impacted this space?

Topic 5 - If I understand this correctly, organizations would put generative AI on a problem to describe a need for a data set, the model would then evaluate the available data and create a quality synthetic or “fake” dataset. How would the organization verify the quality of the dataset? How would they validate that a synthetic data set is as good as the original data?

Topic 6 - Let’s talk about resources for a bit. When I think of GenAI and training, I think of large amounts of hardware and in particular GPU’s that might have limited availability. Is that true here? Also, is this on-prem or in the cloud, or both? 


FEEDBACK?


Email: show at the cloudcast dot netTwitter: @cloudcastpodInstagram: @cloudcastpodTikTok: @cloudcastpod

With the AI Era upon us, the challenge of trying to learn and make sense of the technologies, the business opportunities and the pitfalls is both exciting and equally terrifying.  
SHOW: 812
SHOW TRANSCRIPT: The Cloudcast #812
SHOW VIDEO: https://youtube.com/@TheCloudcastNET 
CLOUD NEWS OF THE WEEK - http://bit.ly/cloudcast-cnotw
CHECK OUT OUR NEW PODCAST - "CLOUDCAST BASICS"


SHOW NOTES:


WHEN WAS THE LAST TIME YOU REALLY LEARNED SOMETHING NEW IN TECH?
Cloud was all new, but not completelyThe pace of cloud innovation seemed fast, but not in comparison to AIThe focus was around one cloud vs. a new source every weekPICK A TOPIC, READ, BE CONFUSED, LOOK FOR CONFIRMATION, RINSE, REPEAT
Where to start the learning process? Where to pick a source of learning?(Coursera, MIT online, blogs/videos/search, etc.)How to determine legitimacy of the sources? How far to learn before stopping to make sense? Trying to relate it to something you already know? How and when to ask questions? FEEDBACK?
Email: show at the cloudcast dot netTwitter: @cloudcastpodInstagram: @cloudcastpodTikTok: @cloudcastpod

Brad Winett (President/Co-founder @TrackItCloud) talks about platforms for entertainment and media. Topics include use cases, partnering with AWS, and creation and consulting services. We even dig into AR and VR a bit at the end.
SHOW: 811

CLOUD NEWS OF THE WEEK - http://bit.ly/cloudcast-cnotw

NEW TO CLOUD? CHECK OUT OUR OTHER PODCAST - "CLOUDCAST BASICS"

SHOW NOTES:
TrackIt (homepage)Topic 1 - Our topic for today is media and entertainment in the cloud. I don’t believe we have ever done a show specifically on this topic, and there are some considerations worth talking about. For today, we have Brad Winett, President and Co-founder at TrackIt. Brad, welcome to the show. Let’s jump right in. The media industry as a whole has undergone major change, just like many others. Most of us see it from the consumer end as a cord-cutter. What made you jump into this market and this industry specifically?

Topic 2 - Platforms and content distribution in the early days of cloud was a differentiator. I think back to Netflix, they initially had a market advantage because they were able to scale better and to more devices than anyone and even open sourced a number of internally developed items and were the AWS poster child. Over time, these user experiences have become the norm. How should people out there think about media platforms? Are we past the days of build your own?

Topic 3 - What about use cases? Media streaming is pretty broad. What does a normal customer look like? Is this big streaming services, smaller companies, etc?

Topic 4 - How much of the tech stack is AWS products and how much of the stack is custom typically? Walk us through what a media streaming stack looks like. How is this different from a SaaS provider providing a turnkey service?

Topic 5 - I know TrackIt is a big AWS partner. Give everyone an overview of the landscape of AWS Partnership these days. Do you provide mainly professional services and consulting?

Topic 6 - Where does open-source software fit into this?

Topic 7 - I feel the standard last question these days is how AI will potentially enhance or impact this is some way.

FEEDBACK?
Email: show at the cloudcast dot netTwitter: @cloudcastpodInstagram: @cloudcastpodTikTok: @cloudcastpod

What will be the adoption patterns for AI within the Enterprise? Will it follow the early days of Cloud Computing, or will new and different patterns emerge? 
SHOW: 810
SHOW TRANSCRIPT: Cloudcast #810 
SHOW VIDEO: https://youtube.com/@TheCloudcastNET 
CLOUD NEWS OF THE WEEK - http://bit.ly/cloudcast-cnotw
CHECK OUT OUR NEW PODCAST - "CLOUDCAST BASICS"


SHOW NOTES:
WHAT WERE THE PATTERNS FOR ENTERPRISE IT AND CLOUD?
Shadow ITHigh-scalability or Short-term Projects (and experimentation)Migration via “Cloud First” initiativesDifficult stuff came lastWHAT’S DIFFERENT ABOUT AI vs. CLOUD?
CPU to CPU was easier to calculate vs. CPU + GPUHave we learned any lessons about how to value people's productivity?Does Enterprise AI need a Crawl, Walk, Run scenario? Do they need to be sequential and linked? Are Enterprise AI use-cases well defined? How long is the Enterprise willing to fail at experiments? What’s the Enterprise tolerance for GenAI “flaws” (e.g. hallucinations, lack of citations, etc.)Will GenAI rejuvenate Predictive AI projects in the Enterprise? 

FEEDBACK?
Email: show at the cloudcast dot netTwitter: @cloudcastpodInstagram: @cloudcastpodTikTok: @cloudcastpod