Subscribe to Newsletter

Is a Pre-Commercial Procurement tender suitable for procuring cloud services? Read the views of the HNSciCloud winning consortia and learn more on their innovative solutions

Last November four winning consortia were awarded the HNSciCloud Pre-Commercial Procurement (PCP) tender. These Consortia are now developing the designs of the resulting European hybrid cloud platform which will be delivered at the end of February. The three most convincing designs will go through to the Prototyping Phase.
As the Pre-Commercial Procurement instrument is a rather new way of procuring cloud services in the European landscape, we asked the representatives of the winning consortia to share with us the pros and the cons of their bidding experience.

T-System’s Account Director Jurry de la Mar, IBM’s researcher and cloud architect Ezra Silvera, the SixSq CEO Marc-Elian Begin and Indra Sistemas’ Director Alfonso Rios provided their views on this topic.

Almost all of them agree that bidding for the tender was really a complex task. This was partially due to the challenging tender requirements (such as the handling of large volumes of data, networks, and security to make the science applications work in a cloud environment) and also because of the novelty of the PCP instrument with which none of the bidders were familiar.

However, they all agreed that the support they received from the HNSciCloud buyers was excellent and fundamental in guiding them through a successful proposal. The Buyers Group was very responsive to questions and made sure all vendors received the requested clarifications. The Open Market Consultation workshop held in March and the Information Session held in September were particularly important: they helped build awareness and facilitate mutual understanding.

This continuous open dialogue between demand and supply side allowed the bidders to start the design phase with a lot more maturity with respect to what usually happens with “traditional” tenders. This factor is especially important for SMEs that usually invest a lot of time and resources in applying for bids. With the PCP, the time and the money invested upfront by the SMEs can be considered all capitalised for the design phase.

The dialogue is indeed also the very critical aspect of the PCP because vendors and buyers need to gain insight about the requirements and the framing conditions on each other's side. They have to evolve the requirements and the solution on this basis. This needs to cover all aspects, i.e., technical, commercial and legal ones (to be adequately reflected in the contract).

Overall, all the bidders are unanimous in affirming that the Pre-Commercial Procurement is a very powerful instrument because it gives the possibility to develop proof of concept, covering part of the cost, with the guarantee for the buyer to benefit from R&D and innovation. It’s obviously still a new instrument, particularly in the field of cloud products and services, which  industry is still not used to.

After having analysed the bidding phase, we asked the consortia to summarise what are the strongest points of the solution they are going to design.

Jurry de la Mar, Account Director in the Public Sector for T-Systems (involved in the T-Systems-Huawei-Cyfronet-Divia and RHEA Group-T-Systems-exoscale-SixSq consortia), says that with their new public cloud service, the Open Telekom Cloud, which was introduced in March 2016, they already had in place a number of the functionalities that were required in the tender. “Next to this public cloud service, we also have a strong community that supports us, the OpenStack community. We believe this is one of the strongest communities currently in cloud development, and we believe that we can draw a number of innovations from this community and make them available to scientists. Our company, as part of Deutsche Telekom, has a strong legacy of networking. We also have long-term experience working with GEANT, the European academic network, and we believe that is another strong point to give access to large amounts of data and be able to transfer large amounts of data across Europe. Lastly, we have good experience working in the science sector. I always say that R&D is in our DNA; as an ICT provider, we always have the next innovation and evolution of our products in mind. Especially within the science sector, we have been able to develop new versions of products. We believe this tender is also an excellent base from which to continue to do that.”

According to Ezra Silvera, Researcher and Cloud Architect for IBM, the strongest points of the solution within its consortium are: “Our design is built on proven technologies and an existing rich IaaS layer, we aim to present a very stable solution that can scale to large numbers while maintaining flexibility. Secondly, we believe that for the data management (i.e., data access and storage) and container support modules, our design has unique capabilities and will demonstrate strength.”

Alfonso Rios, Director of PaaS Cloud Unit at Indra Sistemas, highlights that the differential point of its consortium (Indra, HPE, Advania, SixSq) is that they want to enable a platform for a wide community of providers. This community enrolls different types of companies. Some of them are big, some are small or medium-sized, and they have many different types of cloud platforms.

Finally what do SMEs part of the consortia say? Marc-Elian Begin, CEO and Co-Founder of SixSq, part of the T-Systems-Huawei-Cyfronet-Divia and RHEA Group-T-Systems-exoscale-SixSq consortia affirms that what they have tried to bring in terms of innovation and unique selling propositions to the two bids, is really the idea of embracing the hybrid and multi-cloud right from the start and providing an open system that users can learn from and evolve with. “It’s really part of setting the users on a transformation journey that takes advantage of the cloud as it is right now, but also able to plan future improvements to influence their roadmap so they can take full advantage of what the cloud can do. It’s multi-cloud in its hybrid form that we believe is the future”.

All the premises for the successful implementation of the design phase seem in place. Are you interested with the results of the Design Phases? Stay tuned, the designs that will enter the Prototype phase will be announced in early April 2017.

 

THE FULL INTERVIEWS ARE AVAILABLE BELOW

Do you have any feedback regarding the Tender Application?

 “It was certainly not a normal tender as we see them. The major challenge was to address the requirements, especially the handing of large volumes of data, networks, and security to make the science applications work in a cloud environment“ Jurry de la Mar, Account Director in the Public Sector for T-Systems.

Ezra Silvera, Researcher and Cloud Architect for IBM (Consortium 2): Against this background, the request for proposals was prepared very professionally. The Buyers Group was very responsive to questions and made sure all vendors received the requested clarifications. The workshops held in March and in September were particularly important. They helped build awareness and facilitated mutual understanding.

Marc-Elian Begin, CEO and Co-Founder of SixSq (Consortium 3): The Tender Application for HNSciCloud has been a process in which SixSq was involved from the early phases. It’s been interesting for us to be able to understand what the buyer’s group were looking for and contribute to their thinking. When we ended up with a winning bid, there was a lot of maturity already put into the process, which was obvious for us. It meant that it was not necessarily trivial but reasonably straightforward to respond to the tender, which made our life simpler. As an SME, responding to a complex bid is always a tradeoff we have to make with the probability of winning. So, in this case it was a very positive result based on the investment we put into the bid.

Alfonso Rios, Director of PaaS Cloud Unit at Indra Sistemas (Consortium 4): It has been a complex tender application, but we appreciate all the effort the community has put into presenting a pretty solid list of criteria. They have helped us with every issue we encountered along the way, and in general we are quite happy with it.

What do you think about the Pre-Commercial Procurement Instrument?

Jurry de la Mar, Account Director in the Public Sector for T-Systems (Consortium 1): We believe it’s a very good instrument; it sets a very good governance in place for achieving R&D and innovation. It’s obviously still new to this environment, particularly in developing cloud products and services. Therefore, it is a new challenge for us to see how we can work with PCP. However, judging from what we have seen so far, everything is there to make this possible over the next 18 months of the project.

Ezra Silvera, Researcher and Cloud Architect for IBM (Consortium 2): For first-of-a-kind solutions, pre-commercial procurement provides the public sector with a very suitable procurement procedure. The ability for interaction, which is provided by the pre-commercial procurement, is critical. The vendors and the buyers need to gain insight about the requirements and the framing conditions on each other's side. They have to evolve the requirements and the solution on this basis. This needs to involve all aspects, i.e., technical, commercial and legal ones (to be adequately reflected in the contract).

Marc-Elian Begin, CEO and Co-Founder of SixSq (Consortium 3): The PCP instrument is an interesting instrument. It’s a great way for us to work with great partners and great organisations, in the form of the buyers group in this case. Having said that, there’s a number of constraints that are difficult to work with: the staged phases and especially the payment system creates challenges for us. As an SME, cash is king, so it can be difficult to put into place all the innovation we wanted to because of those constraints. But overall the idea of engaging early to address a problem that the buyers group wants to address in such a way is a great thing. I just think that there’s a number of improvements that will make the process even more attractive for SMEs like us.

Alfonso Rios, Director of PaaS Cloud Unit at Indra Sistemas (Consortium 4): As an instrument, it’s great because it will allow us as companies to develop proof of concept, covering part of the cost (hopefully most of the cost). However, it’s still a new instrument; we have to get a little more experience with it. But I think it’s great idea in order to bridge the gap that normally exists between the public sector and the private sector in order to learn from both sides how we can get closer to achieving our common interests.

In your opinion, what are the strongest points of the solution that you are going to design for the upcoming phase?

Jurry de la Mar, Account Director in the Public Sector for T-Systems (Consortium 1): Starting from the requirements, we identified that, with our new public cloud service, we have a very good starting point: the Open Telekom Cloud, which was introduced in March 2016, already brings a number of the functionalities that have been required. We have also been able to expand the Open Telekom Cloud with new functionality that is exactly within the requirements: things like container management and handling of large-scale data. Next to this public cloud service, we have a strong community that supports us. The Open Telekom Cloud is based on OpenStack. We believe this is one of the strongest communities currently in cloud development, and we believe that we can draw a number of innovations from this community to make available to the scientists. Our company, as part of Deutsche Telekom, has a strong legacy of networking. We also have long-term experience working with GÉANT, the European academic network, and we believe that is another strong point to give access to large amounts of data and be able to transfer large amounts of data across Europe. Lastly, we have good experience working in the science sector. I always say that R&D is in our DNA; as an ICT provider, we always have the next innovation and evolution of our products in mind. Especially within the science sector, we have been able to develop new versions of products. We believe this tender is also an excellent base from which to continue to do that.

Ezra Silvera, Researcher and Cloud Architect for IBM (Consortium 2): There are a few areas in which we believe our design will demonstrate strength. First, because our design is built on proven technologies and an existing rich IaaS layer, we aim to present a very stable solution that can scale to large numbers while maintaining flexibility. Second, we believe that for the data management (i.e., data access and storage) and container support modules, our design has unique capabilities and will demonstrate strength.

Marc-Elian Begin, CEO and Co-Founder of SixSq (Consortium 3): What we’ve tried to bring in terms of innovation and unique selling propositions to the two bids, is really the idea of embracing the hybrid and multi-cloud right from the start and providing an open system that users can learn from and evolve with. It’s really part of setting the users on a transformation journey to take advantage of the cloud as it is right now, but also able to plan future improvements to influence their roadmap so they can take full advantage of what the cloud can do. It’s multi-cloud in its hybrid form that we believe is the future.

Alfonso Rios, Director of PaaS Cloud Unit at Indra Sistemas (Consortium 4): For the upcoming phase, one differential point in our approach is that we want to enable a platform for a wide community of providers. This community enrolls different types of companies. Some of them are big, some are small or medium, and they have many different types of cloud platforms. Therefore, we have to provide them with an open platform that would enable them to become providers of the open science community. This is the main differential approach from us.

Who's behind HNSciCloud?

HNSciCloud Pre-Commercial Procurement is driven by ten leading research organisations from across Europe.