>> Hello everyone. Welcome to theCUBEs coverage of Compute Engineered for your Hybrid World sponsored by HPE and Intel. Today we're going to dive into advanced performance of VDI with the fourth gen Intel Zion scalable processors. Hello I'm John Furrier, the host of theCUBE. My guests today are Alan Chu, Director of Data Center Performance and Competition for Intel as well as Denis Kondakov who's the VDI product manager at HPE, and also joining us is Cynthia Sustiva, CAD/CAM product manager at HPE. Thanks for coming on, really appreciate you guys taking the time. >> Thank you. >> So accelerating VDI to the Edge. That's the topic of this topic here today. Let's get into it, Dennis, tell us about the new HPE ProLiant DL321 Gen 11 server. >> Okay, absolutely. Hello everybody. So HP ProLiant DL320 Gen 11 server is the new age center CCO and density optimized compact server, compact form factor server. It enables to modernize and power at the next generation of workloads in the diverse rec environment at the Edge in an industry standard designed with flexible scale for advanced graphics and compute. So it is one unit, one processor rec optimized server that can be deployed in the enterprise data center as well as at the remote office at end age. >> Cynthia HPE has announced another server, the ProLiant ML350. What can you tell us about that? >> Yeah, so the HPE ProLiant ML350 Gen 11 server is a powerful tower solution for a wide range of workloads. It is ideal for remote office compute with NextGen performance and expandability with two processors in tower form factor. This enables the server to be used not only in the data center environment, but also in the open office space as a powerful workstation use case. >> Dennis mentioned both servers are empowered by the fourth gen Intel Zion scale of process. Can you talk about the relationship between Intel HPE to get this done? How do you guys come together, what's behind the scenes? Share as much as you can. >> Yeah, thanks a lot John. So without a doubt it takes a lot to put all this together and I think the partnership that HPE and Intel bring together is a little bit of a critical point for us to be able to deliver to our customers. And I'm really thrilled to say that these leading Edge solutions that Dennis and Cynthia just talked about, they're built on the foundation of our fourth Gen Z on scalable platform that's trying to meet a wide variety of deployments for today and into the future. So I think the key point of it is we're together trying to drive leading performance with built-in acceleration and in order to deliver a lot of the business values to our customers, both HP and Intels, look to scale, drive down costs and deliver new services. >> You got the fourth Gen Z on, you got the Gen 11 and multiple ProLiants, a lot of action going on. Again, I love when these next gens come out. Can each of you guys comment and share what are the use cases for each of the systems? Because I think what we're looking at here is the next level innovation. What are some of the use cases on the systems? >> Yeah, so for the ML350, in the modern world where more and more data are generated at the Edge, we need to deploy computer infrastructure where the data is generated. So smaller form factor service will satisfy the requirements of S&B customers or remote and branch offices to deliver required performance redundancy where we're needed. This type of locations can be lacking dedicated facilities with strict humidity, temperature and noise isolation control. The server, the ML350 Gen 11 can be used as a powerful workstation sitting under a desk in the office or open space as well as the server for visualized workloads. It is a productivity workhorse with the ability to scale and adapt to any environment. One of the use cases can be for hosting digital workplace for manufacturing CAD/CAM engineering or oil and gas customers industry. So this server can be used as a high end bare metal workstation for local end users or it can be virtualized desktop solution environments for local and remote users. And talk about the DL320 Gen 11, I will pass it on to Dennis. >> Okay. >> Sure. So when we are talking about age of location we are talking about very specific requirements. So we need to provide solution building blocks that will empower and performance efficient, secure available for scaling up and down in a smaller increments than compared to the enterprise data center and of course redundant. So DL 320 Gen 11 server is the perfect server to satisfy all of those requirements. So for example, S&B customers can build a video solution, for example starting with just two HP ProLiant TL320 Gen 11 servers that will provide sufficient performance for high density video solution and at the same time be redundant and enable it for scaling up as required. So for VGI use cases it can be used for high density general VDI without GP acceleration or for a high performance VDI with virtual VGPU. So thanks to the modern modular architecture that is used on the server, it can be tailored for GPU or high density storage deployment with software defined compute and storage environment and to provide greater details on your Intel view I'm going to pass to Alan. >> Thanks a lot Dennis and I loved how you're both seeing the importance of how we scale and the applicability of the use cases of both the ML350 and DL320 solutions. So scalability is certainly a key tenant towards how we're delivering Intel's Zion scalable platform. It is called Zion scalable after all. And we know that deployments are happening in all different sorts of environments. And I think Cynthia you talked a little bit about kind of a environmental factors that go into how we're designing and I think a lot of people think of a traditional data center with all the bells and whistles and cooling technology where it sometimes might just be a dusty closet in the Edge. So we're defining fortunes you see on scalable to kind of tackle all those different environments and keep that in mind. Our SKUs range from low to high power, general purpose to segment optimize. We're supporting long life use cases so that all goes into account in delivering value to our customers. A lot of the latency sensitive nature of these Edge deployments also benefit greatly from monolithic architectures. And with our latest CPUs we do maintain quite a bit of that with many of our SKUs and delivering higher frequencies along with those SKUs optimized for those specific workloads in networking. So in the end we're looking to drive scalability. We're looking to drive value in a lot of our end users most important KPIs, whether it's latency throughput or efficiency and 4th Gen Z on scalable is looking to deliver that with 60 cores up to 60 cores, the most builtin accelerators of any CPUs in the market. And really the true technology transitions of the platform with DDR5, PCIE, Gen five and CXL. >> Love the scalability story, love the performance. We're going to take a break. Thanks Cynthia, Dennis. Now we're going to come back on our next segment after a quick break to discuss the performance and the benefits of the fourth Gen Intel Zion Scalable. You're watching theCUBE, the leader in high tech coverage, be right back. Welcome back around. We're continuing theCUBE's coverage of compute engineer for your hybrid world. I'm John Furrier, I'm joined by Alan Chu from Intel and Denis Konikoff and Cynthia Sistia from HPE. Welcome back. Cynthia, let's start with you. Can you tell us the benefits of the fourth Gen Intel Zion scale process for the HP Gen 11 server? >> Yeah, so HP ProLiant Gen 11 servers support DDR five memory which delivers increased bandwidth and lower power consumption. There are 32 DDR five dim slots with up to eight terabyte total on ML350 and 16 DDR five dim slots with up to two terabytes total on DL320. So we deliver more memory at a greater bandwidth. Also PCIE 5.0 delivers an increased bandwidth and greater number of lanes. So when we say increased number of lanes we need to remember that each lane delivers more bandwidth than lanes of the previous generation plus. Also a flexible storage configuration on HPDO 320 Gen 11 makes it an ideal server for establishing software defined compute and storage solution at the Edge. When we consider a server for VDI workloads, we need to keep the right balance between the number of cords and CPU frequency in order to deliver the desire environment density and noncompromised user experience. So the new server generation supports a greater number of single wide and global wide GPU use to deliver more graphic accelerated virtual desktops per server unit than ever before. HPE ProLiant ML 350 Gen 11 server supports up to four double wide GPUs or up to eight single wide GPUs. When the signing GPU accelerated solutions the number of GPUs available in the system and consistently the number of BGPUs that can be provisioned for VMs in the binding factor rather than CPU course or memory. So HPE ProLiant Gen 11 servers with Intel fourth generation science scalable processors enable us to deliver more virtual desktops per server than ever before. And with that I will pass it on to Alan to provide more details on the new Gen CPU performance. >> Thanks Cynthia. So you brought up I think a really great point earlier about the importance of achieving the right balance. So between the both of us, Intel and HPE, I'm sure we've heard countless feedback about how we should be optimizing efficiency for our customers and with four Gen Z and scalable in HP ProLiant Gen 11 servers I think we achieved just that with our built-in accelerator. So built-in acceleration delivers not only the revolutionary performance, but enables significant offload from valuable core execution. That offload unlocks a lot of previously unrealized execution efficiency. So for example, with quick assist technology built in, running engine X, TLS encryption to drive 65,000 connections per second we can offload up to 47% of the course that do other work. Accelerating AI inferences with AMX, that's 10X higher performance and we're now unlocking realtime inferencing. It's becoming an element in every workload from the data center to the Edge. And lastly, so with faster and more efficient database performance with RocksDB, we're executing with Intel in-memory analytics accelerator we're able to deliver 2X the performance per watt than prior gen. So I'll say it's that kind of offload that is really going to enable more and more virtualized desktops or users for any given deployment. >> Thanks everyone. We still got a lot more to discuss with Cynthia, Dennis and Allen, but we're going to take a break. Quick break before wrapping things up. You're watching theCUBE, the leader in tech coverage. We'll be right back. Okay, welcome back everyone to theCUBEs coverage of Compute Engineered for your Hybrid World. I'm John Furrier. We'll be wrapping up our discussion on advanced performance of VDI with the fourth gen Intel Zion scalable processers. Welcome back everyone. Dennis, we'll start with you. Let's continue our conversation and turn our attention to security. Obviously security is baked in from day zero as they say. What are some of the new security features or the key security features for the HP ProLiant Gen 11 server? >> Sure, I would like to start with the balance, right? We were talking about performance, we were talking about density, but Alan mentioned about the balance. So what about the security? The security is really important aspect especially if we're talking about solutions deployed at the H. When the security is not active but other aspects of the environment become non-important. And HP is uniquely positioned to deliver the best in class security solution on the market starting with the trusted supply chain and factories and silicon route of trust implemented from the factory. So the new ISO6 supports added protection leveraging SPDM for component authorization and not only enabled for the embedded server management, but also it is integrated with HP GreenLake compute ops manager that enables environment for secure and optimized configuration deployment and even lifecycle management starting from the single server deployed on the Edge and all the way up to the full scale distributed data center. So it brings uncompromised and trusted solution to customers fully protected at all tiers, hardware, firmware, hypervisor, operational system application and data. And the new intel CPUs play an important role in the securing of the platform. So Alan- >> Yeah, thanks. So Intel, I think our zero trust strategy toward security is a really great and a really strong parallel to all the focus that HPE is also bringing to that segment and market. We have even invested in a lot of hardware enabled security technologies like SGX designed to enhance data protection at rest in motion and in use. SGX'S application isolation is the most deployed, researched and battle tested confidential computing technology for the data center market and with the smallest trust boundary of any solution in market. So as we've talked about a little bit about virtualized use cases a lot of virtualized applications rely also on encryption whether bulk or specific ciphers. And this is again an area where we've seen the opportunity for offload to Intel's quick assist technology to encrypt within a single data flow. I think Intel and HP together, we are really providing security at all facets of execution today. >> I love that Software Guard Extension, SGX, also silicon root of trust. We've heard a lot about great stuff. Congratulations, security's very critical as we see more and more. Got to be embedded, got to be completely zero trust. Final question for you guys. Can you share any messages you'd like to share with the audience each of you, what should they walk away from this? What's in it for them? What does all this mean? >> Yeah, so I'll start. Yes, so to wrap it up, HPR Proliant Gen 11 servers are built on four generation science scalable processors to enable high density and extreme performance with high performance CDR five memory and PCI 5.0 plus HP engine engineered and validated workload solutions provide better ROI in any consumption model and prefer by a customer from Edge to Cloud. >> Dennis? >> And yeah, so you are talking about all of the great features that the new generation servers are bringing to our customers, but at the same time, customer IT organization should be ready to enable, configure, support, and fine tune all of these great features for the new server generation. And this is not an obvious task. It requires investments, skills, knowledge and experience. And HP is ready to step up and help customers at any desired skill with the HP Greenlake H2 cloud platform that enables customers for cloud like experience and convenience and the flexibility with the security of the infrastructure deployed in the private data center or in the Edge. So while consuming all of the HP solutions, customer have flexibility to choose the right level of the service delivered from HP GreenLake, starting from hardwares as a service and scale up or down is required to consume the full stack of the hardwares and software as a service with an option to paper use. >> Awesome. Alan, final word. >> Yeah. What should we walk away with? >> Yeah, thanks. So I'd say that we've talked a lot about the systems here in question with HP ProLiant Gen 11 and they're delivering on a lot of the business outcomes that our customers require in order to optimize for operational efficiency or to optimize for just to, well maybe just to enable what they want to do in, with their customers enabling new features, enabling new capabilities. Underpinning all of that is our fourth Gen Zion scalable platform. Whether it's the technology transitions that we're driving with DDR5 PCIA Gen 5 or the raw performance efficiency and scalability of the platform in CPU, I think we're here for our customers in delivering to it. >> That's great stuff. Alan, Dennis, Cynthia, thank you so much for taking the time to do a deep dive in the advanced performance of VDI with the fourth Gen Intel Zion scalable process. And congratulations on Gen 11 ProLiant. You get some great servers there and again next Gen's here. Thanks for taking the time. >> Thank you so much for having us here. >> Okay, this is theCUBEs keeps coverage of Compute Engineered for your Hybrid World sponsored by HP and Intel. I'm John Furrier for theCUBE. Accelerate VDI at the Edge. Thanks for watching.

(upbeat digital music) >> Hey welcome back, get ready. Jeff Frick here with theCUBE. We're in our Palo Alto studio for a Cube Conversation. Again, we love talking with little companies, emerging companies, kind of maybe technology you haven't heard of before and we're excited to have our next guest 'cause he's right in the heart of security space, which is always a hot topic, continues to be a hot topic and will never go away 'cause the bad guys they just keep working hard to try to break everything that we create. So our next guest is Ambuj Kumar, the co-founder and CEO of Fortanix. Ambuj welcome. >> Thank you, Jeff. >> So give, for the people who aren't familiar with Fortanix kind of the basic 101. >> Yeah, so if you look at all the security today, it falls into three categories. One is protecting your data address. So what that means is, if somebody steals your laptop, how do you protect your hard drive from getting exposed? >> Right. >> So we use encryption for that. Similarly, we also use encryption to secure our data in use. So we connect to some bank website and our data goes encrypted through TELUS and so what that means is if somebody's doing wiretapping our data is protected. However once the applications start to run, whether it's in your data center or public cloud, then the data applications are being exposed. So to fix that Runtime vulnerabilities what the industry has done so far is to secure the infrastructure, try to secure the infrastructure and that is $80 billion per year industry. But we have failed to that because infrastructure is just so vastly complex. So what we do is we use something called Runtime encryption and idea is that your data and applications remain encrypted, so even when people who are running your cloud they're untrusted and they want to get your data, they can't do anything with it. >> So, a lot of stuff there to unpack. So first off we know the perimeter systems don't work anymore. >> Yeah >> I mean you got to put them up they do some level of stuff But you can't secure the perimeter anymore. So it is all this kind of working your security >> Yeah and the encryption all the way through the process. But this is pretty interesting I've never heard of encryption actually at Runtime, I mean it begs the question, you know how does the microprocessor run the encrypted data? >> That's right So it's a long research problem in security. People had been working on something called Fully homomorphic encryption and the idea is that: Can I take my program encrypted data encrypted and run in totally untrusted environment and give you the result that you can decrypt. Chances are that you can do that with very simple programs, like if you're adding some numbers, multiplying those numbers and even in those cases slow by many orders of magnitude. So what normally some operations takes one second will it will take three years. >> Okay >> Not good. >> Laughs >> So what we do is we use some new instructions from Intel called Software Guard Extension, Intel SGX and your data and your programs, they get decrypted in a secure region of CPU So all the memory, all the operating systems accessible things, anything that can be touched by any other process, they only can look at encrypted stuff. Your data get decrypted right when instructions are working on them and at that point it is accessible only to your write process. >> Right. >> So you use this hardware capability to accelerate the encryption decryption. So we can provide all the benefits of fully owned morphic encryption at a performance that is totally acceptable to our customers. >> So let me make sure I understand, So it decrypts it literally at the last possible obviously not second >> Yeah but last possible (laughs) in microprocessor time >> Yeah cycle, runs that process and then is write only to the output of that process. And is that immediately encrypted again >> Exactly >> On the write side as well? >> Yeah Yeah, exactly. Exactly. >> (laughs) So you mentioned the Intel instructions So is this relatively new, the SGX? >> Yeah, so we were first vendor to commercialize Intel SGX, its a new technology, but it's coming in all their CPU's so right now it's in all client CPU's, and some of the data centers CPU's But five years from now all the CPU's you will get from Intel will hopefully have this technology >> Right So obviously Skylake >> Yeah Skylake has it and all newer architecture. >> Wow So a little bit more about the company How long you guys been around, how long you been working on this problem you know funding kind of give us the overview on the company. >> Yeah >> So I have been working on encryption for last seven years the company was founded two years ago >> Okay >> We are funded by some well known security VC's including Foundation Capital and NeoTribe Ventures >> Okay >> We are widely recognized as the pioneers in this field that we are creating Runtime encryption. Recognized by Gartner's Cool Vendor we came number two in RSA Innovation Sandbox you know hundreds of security companies. We have several S&P 500 customers already so we are deployed in their products and environment, we are securing trillions of dollars of assets in realtime. Our goal is to convince CIA to run their most prestigious most sensitive applications on some untrusted cloud in some enemy country. >> Laughs >> It's a long shot >> Are you doing like a POC of something like that with them? Are you in active conversations or is that more of kind of a philosophical goal? >> I cannot confirm of deny that >> Okay, fair enough >> But that's our goal. And until we achieve that, we have something to keep working on. >> Okay. And then where do you guys sit kind of in the world of public clouds with AWS and Azure and Google versus either private (mumbles) or multiple clouds inside the company or you know some of these other kind of options like we hear like the Equinix which I think is one of the places >> Yeah >> How's that work? >> Yeah So our goal is to extricate security from infrastructure So in the end, our goal is that infrastructure will provide you compute cycles and the security will come from the customers, end customers who are developing the applications and deploying the applications. >> Right >> So its cloud agnostic security so meaning that we will go after on-prem customers, we'll go after public cloud, colo and all of that >> Right >> So in the meantime for our go-to market what we did was we partnered with two of really well known strong forces in the industry, one is IBM Cloud >> Yeah where IBM is putting this servers and running our technology and with Equinix, which is world's largest data provider and so if you are in any of the public cloud, if you are in IBM cloud you get our security by default so you are continuous running encryption >> Right >> Isolated from all the threats that might be there, or if you are in some other public cloud you can use it Equinix colo so if you have some applications that you don't want to be hacked you can use our SAS service to run those applications encrypted. >> Right And of course Equinix has got the direct connect to all the public clouds >> Yeah >> So minimum latency integration >> Couple of milliseconds. >> with all the other stuff >> in the public cloud. >> Yeah exactly. So what's the expense, both kind of the overhead expense on the computing side to do this when it's done properly and then what's the expense to run this is this something that is expensive can only be used for the most critical applications, or do you see this several times being more general purpose execution? >> So its will be used to secure anything that you don't want to be hacked and the cost of using Runtime encryption is minimal so I expect it to be wisely adopted and we make it really easy for developers and security organizations to use this technology. So you have to bring in your container and then Fortanix process attaches to your container you don't need to recompile your source code we never get to look at your source code there's no binary transfers nothing like that. And then so it's a simple millisecond long process and we give you modified container and now you can take this modified container run on any cloud you want and if it runs it runs securely. From that point onwards. >> Right And today you just have to make sure its got right microprocessor >> Yeah and in the future hopefully that will be more general purpose. >> Yeah >> Alright So what's next? What are you working on, what's a priority for the balance of 2018? >> Yeah, so we have lots of integration work going on VIA World is coming next week We have support for something called Kermit that allows you to secure your estorial box v send et cetera with Fortanix. Now we are also running integration with some data bases some multi party computers and things like that. So our goal is to make our technology more widely available to a large variety of customers. >> Alight, well Ambuj very interesting story, Encryption at Runtime so >> Yeah >> So we look forward to watching the story unfold. >> Awesome, yeah This is a decade long journey and I think when we have done infrastructure security will be irrelevant. So its going to be very exciting for all the parties involved. >> Alright, we'll keep eye, thanks for stopping by. >> Thanks >> Alrighty, Ambuj Kumar You're watching theCube from our Palo Alto studios See you next time. And thanks for watching. (epic orchestra music)