AWS re: Invent 2019 with Corey Quinn: the biggest announcements of the conference
Corey Quinn, cloud economist at The Duckbill Group, shares his vision on product announcements and pricing strategies revealed at AWS re: Invent 2019.
AWS re: Invent 2019 with Corey Quinn: the biggest announcements of the conference
Corey Quinn, cloud economist at the Duckbill Group, shares his vision on product announcements and pricing strategies revealed at AWS re: Invent 2019.
James Sanders from TechRepublic spoke with Corey Quinn, cloud economist at The Duckbill Group, at AWS re: Invent. Quinn has built a reputation for hot takes on AWS service offerings and billing practices, and his distinct nature has made him a fixed value on Twitter. The following is an edited transcript of the conversation.
Corey Quinn: From my perspective, Compute Optimizer is one of the most interesting things that emerged based on what I do. It uses machine learning and instead of applying it to momentary things that can find everything in a stack of data, except for a business model, it applies it to a limited problem domain of the AWS invoice. Are you using copies of the correct size? And if not, this is what we recommend instead. It is valuable and useful for companies trying to enter their workload.
Graviton2 is interesting and I think we will see the rise of ARM in the somewhat near future. People are going to do some analysis and figure out what it takes to transfer our workload from x86 to ARM, and the answer is just moving – it already works. It also appears that a little more work needs to be done. Cross-compiling is not just about using GCC if you are angry.
SEE: AWS Graviton2: What it means for Arm in the data center, cloud, enterprise, AWS (ZDNet)
CPU performance and cost savings
The interesting thing is that for comparable or allegedly superior performance, they save around 20% in costs. Don’t quote me about that. I don’t remember facts and figures like some people do. The counter argument is now . so the moment they begin to get that level of access to discounting, companies are going to make the switch themselves. It feels like we’re on the eve of a revolution in the direction of ARM, and what is likely to start tilting is when Apple releases a laptop with an ARM chip in it, and to be honest, the blame lies largely with Intel. A lot of blown deadlines, a lot of overloading and under-performance have really led to a point where many manufacturers seem pretty disgusted. Now Intel has a stand here: Invent Expo Hall, but I can’t figure out why for life.
SEE: The awful, bad, awful week of Intel amid AWS moves Qualcomm’s Arm (ZDNet)
That is going to be so exciting, because we see future generations of instances coming online in all different processor manufacturer families. What are we going to see in terms of performance? And for good reason, no matter how much we like to talk about how critical CPU performance is, you look at the traditional workload in most companies and the computers are incredibly bored. It’s like looking at an IBM keynote.
AWS outpost and hardware suppliers
James Sanders: What impact does AWS Outpost have on traditional hardware suppliers, such as HPE and Dell?
Corey Quinn: I do not necessarily know how to answer that question because I am not inclined to deal with companies with huge on-premise environments, only based on the nature of what I do, or if they have it, they are quick search for evacuation for various reasons. I don’t think those companies will dry up quickly and blow away, but I think there is a clear long-term story for them. If you start today with a Fortune 500 that grows from a small, sloppy start-up to that gigantic company, where are you heading to a number of traditional IT suppliers for companies? The unfortunate truth is that, as things stand now, you will never be their customers. They need to find a way to bridge that gap in a compelling and authentic way, and at the moment I don’t hear a great route map to get there, articulated by a major player in this space.
Amazon Braket is quantum in the cloud
Amazon Braket, or whatever you want to pronounce, or briquettes? The statement is the most interesting about it, because there is now a lot of talk about it and absolutely nothing demonstrative. It is technology of the future that is not present today in any sense in practice or application. When that changes, they want to be able to say that they have had something and have looked at it for a long time, but there is nothing like a paying customer that I can buy today in the world of Quantum that solves a problem I have. And until that changes, we will basically see no significant traction in that market. But given that their competitors say something, they must do the same.
SEE: Amazon Braket’s quantum computing service sticks three unrelated systems together (TechRepublic)
New Amazon products named after women
We generally look at things such as Amazon Macie and Kendra that have just been released. Generally, if Amazon calls a product after a woman, check the price because it is likely to be excessive. I do not know or understand why they chose to walk that path; Kendra is great, but it is incredibly expensive. Macie is interesting, but ridiculously expensive. Again, they do not have a monopoly on expensive and confusing from the perspective of a price model, but it is interesting to see how they charge for new and exciting dimensions.
CodeGuru is also too expensive
The Amazon CodeGuru, which I feel is the answer from a Cloud Guru: “I’m sorry. You called it what?” It is fascinating. It charges 75 cents per month per 100 lines of code review. First stop recording line breaks and use a lot of commas. Secondly, it is particularly awful statistics for developers. And third, even if the price is ultimately reasonable, people suddenly have to think about their code and their development practices in a way that they have not done so far, and that becomes a significant problem. I think it is a bad choice of the price model. I am interested to see how well it works, but today it is ridiculously expensive. I have applications that run with the serverless framework written in JavaScript or Python that cost cents per month that would cost hundreds of dollars to use this service.
But wavelength has potential
Amazon Wavelength is a genius product, and I say this sincerely, because 5G is another buzzword where everyone has to play in this space. Whether it is better or worse, we see that the wireless spectrum is relatively commonplace at this point. Some providers are slightly better than others, but I don’t care which provider I dictate whether or not I can call.
SEE: AWS re: Invent 2019: Why Outposts, Local Zones and Wavelength change the game for companies (TechRepublic)
5G is a buzzword that everyone focuses on. But what makes this fascinating from my perspective is by positioning itself in collaboration with Verizon, Amazon is on the ground floor of the 5G development, giving them access to what’s happening in that space, and their level of investment is really distilling to a set of racks with hardware that they put on a presence point that Verizon owns. It is a great opportunity, from the Amazon perspective.
From the perspective of Verizon, they have the same problem that most telcos seem to have, what the world wants them to be a stupid pipe, but they can only ask so much for it. Well, most companies. Verizon has fascinating ways to get around that. Verizon Math is still one thing, but what’s neat is that they want to offer differentiated services at a higher level, but nobody wants a service that Verizon performs, except the connection itself. So we’ll see if it becomes a competitive differentiator or not. Time will tell, and I will ridicule it anyway.
I will say that Verizon is almost certainly shocked and surprised at how positively the community responds to wavelength. It is the first time in history that Verizon has received a good reception.
SEE: AWS, Verizon want to collaborate on use cases for 5G, cloud and edge computing (ZDNet)
Jassy’s keynote speech
Andy Jassy comes on stage and has an almost insurmountable problem that he has to solve. He asked for an enormous variety of service versions and improvements to be articulated for an incredibly diverse audience coming from a huge variety of industries, backgrounds and problem areas. Every service version is important for someone and no service version is important for everyone.
Making so many different notes with so many different constituencies about so many different things, as well as a three-hour marathon keynote session is incredibly challenging, and I have to say for better or worse, he is more polished every year as a speaker. As someone who comes on stage a bit, mainly because of my constant love affair with the sound of my own voice, I know what it takes to speak for many people. He speaks much more than I usually do.
The AWS house band needs better texts
Oh yes, the house band is depressing, sad and frankly, if Amazon was a friendlier, warmer company that expressed empathy better than they would, they would put the re: Invent house band to sleep. The problem they have is that it is loud; it is very finished. They do song covers very briefly, but they don’t change the lyrics as something that would be humorous or would inspire people to look at things differently, so it just feels a bit hokey. Give me 20 minutes with the text and let’s see what we can come up with instead. If you go for something, go all the way inside.
I think there will always be surprises. Amazon is a company that has not had it for very long or very well, it is a property that I tend to share. The difference really comes down to the fact that I am good at naming things.
Cloud Insights newsletter
Your knowledge base for the latest news on AWS, Microsoft Azure, Google Cloud Platform, Docker, SaaS, IaaS, cloud security, containers, the public cloud, the hybrid cloud, the industry cloud and much more.
Delivered on Mondays
Register today
