Archives

March 16th, 2011 | Tags: , , | Category: Cloud, EDA

Cloud computing: an opportunity for EDA

Cloud computing is becoming more pervasive in many aspects of the day-to-day business of companies: archiving, payroll, CRM, etc.  Whenever the cost of acquiring, maintaining, and scaling one’s own IT resources becomes too high, cloud computing start to become attractive.

ASIC and digital system design is a computing resource-hungry task that would certainly benefit from cloud computing. Still, EDA is mostly staying on the sideline while so many other industries are rapidly shifting towards cloud-based platforms.

Granted, there have been a few incursions of EDA into the cloud.

Synopsys and other EDA vendors have been using Amazon Web Services (AWS) to provide web-based training services. Both Synopsys and Cadence now prefer Xuropa as a web-based training platform, because Xuropa offers a service more tailored to EDA needs (e.g., input/output languages and format).  These services are Infrastructure as a Service (IaaS), i.e., it gives the customer the ability to use processing, network, and storage resources in a flexible manner.

Physware and Plunify are two startups sitting firmly on the cloud-based Software as a Service (SaaS) side. Physware simulates an IC with a true 3D field solver, providing signal integrity, power integrity, and electro-migration analysis. Plunify runs multiple synthesis scenarios to offer a wide area/performance tradeoff to the user. Both use cloud computing to offer a very short, scalable, turn-around-time to their customers. Cadence also has its own cloud-based SaaS offering, with mitigated success so far.

Oh yes, and IBM has been using cloud computing for its own EDA tools for years –with more than 20,000 cores, 150 Tb of memory, running 40,000 jobs per day.

But where does that leave the big three, Synopsys, Cadence, and Mentor, as provider of EDA solutions in the cloud? For instance, physical verification is known to be very tolerant to design partitioning, so a number of physical verification tools can easily take advantage of cloud computing. Logic simulation can also benefit from massive parallelism, even though it is more challenging –partitioning a test bench is trivial, but taking advantage of a design partition for logic simulation is tricky.

It is not like the Big Three do not know what is at stake: unless there is some revolutionary technology in the making, the largest SoC will simply exceed the capacity of today’s synthesis and verification tools. Which means that the semiconductor industry should be eager to access cloud-based SaaS. However there are a few obstacles on the way:

  • Tools needs to be revamped to fit the cloud infrastructure. For verification and simulation, this is not a major bottleneck though.
  • It is unclear what will be the business model of a SaaS EDA in the cloud. But this is a chance to propose new models that would lift the EDA market. Shall we charge by CPU/hour plus bandwidth? Or by the TAT reduction (the more servers, the smaller the TAT, the higher the fee)?
  • Last but not least, security is a major obstacle for design houses to let their IP go into the cloud. But those that express their concerns about security are the same that have a private email in the cloud (Yahoo email, Gmail, Hotmail) and go happily shop on-line. Also data are arguably more reliable in the cloud because of the inherent redundancy required by fault-tolerant platforms.  With time people will come to accept that the cloud is secured enough.

Given that the cloud infrastructure becomes more mainstream thanks to the many open-source resources and accumulated experience in many industries, it is just a matter of time before EDA and its customers are serious about cloud-based SaaS. Let’s just hope that the EDA companies will size the opportunity to reiterate itself as a major enabler of the semiconductor industry, and to propose a new business model that would benefit the industry.

[UPDATE: follow-up post on security and the cloud]

11 comments to Cloud computing: an opportunity for EDA

  • balor123

    Large companies are likely never to become comfortable with storing their data in the cloud. The trend is with time they become more paranoid as technology makes it possible.

    Employees do trust their data in the cloud but that’s really a different issue. Most tech companies consider their IP to be their most valuable asset. Losing control of that IP is equivalent to losing your life savings.

    Also, the amount of data involved in EDA applications is typically pretty large. It’s just inefficient to keep transferring those files back and forth. If you want to make it work, then you’d have to host all the tools used and provide a virtual environment for people to work. Latency working on another computer locally is bad enough imagine somewhere else in the world.

    Last, the cost saved isn’t worthwhile. Most EDA customers tend to prefer to build their own private clouds than use public ones and the cost is insignificant on their balance sheets.

  • Hi Olivier,

    It seems we keep reiterating the same arguments for Cloud for EDA and the naysayers keep reiterating the same arguments against. I can tell you that there are some serious efforts in the works to deploy SaaS in specific niches where it makes sense. In one year, the question will no longer be “if” or “when”, but “how much.”

  • Hi Balor123 (is that your real name? ;)

    Actually, there is a number of large company that moved to the cloud for economic and security reasons. The economics are obvious. As for the security, I argue that having an indelicate employee is as risky as (if not riskier than) a breach in the cloud. The difference is that the cloud has data duplication and data segmentation for large files, which brings reliability and security –breaking into one server or one data center only gives you one piece of the data, not the whole. Also there are a number of very matured encryption techniques that have been in used for decades (think banking…). Try to name one big company that manages its CRM with its own IT resources –and CRM is some sensitive info. A lot of semiconductor companies already share their IPs with service companies: why would the cloud be less secure that sharing bits of the design with a company in India or China?

    The back and forth transfer of large amount of data would be required only a few times if the data is hosted in the cloud. A virtual environment is not that bad if the amount of data that needs to be exchange with the client is small –typically, some HDL code or P&R data for a GUI. Text is trivial. I have been working for years with a RDP server in the US and a laptop in Europe or India, serving me with millions of lines of code and P&R GUI. Think about Google Earth: this looks extremely bandwidth heavy, still you can seamlessly navigate and zoom in/out quickly, because the design of the data layout. Why would that be any more difficult than navigating a GDSII database?

    As for the cost, a lot of EDA customers want to focus on the design, where the added value is, not on the implementation, which is more and more a commodity. Therefore, why would they invest and maintain a costly IT infrastructure when they can pay on-demand? The current trend of some ASIC design houses is to offshore or use service companies for the P&R. A cloud-based SaaS providing a secured platform and delivering a very short TAT looks quite attractive then.

  • Hi Harry,

    Yes, today it’s no longer “if” but “when”. And you may be quite right to say that soon it won’t be “when” but “how much”. True, the semiconductor industry is rather conservative, as is the EDA industry to some extend. So what looks obvious today might take another couple of years before it becomes reality. But the complexity of the design, the cost of implementation and verification, and the pressure of short TAT, all push for EDA proposing cloud-based SaaS to design companies once the later feel more comfortable with the security aspects.

  • It seems to me that it is not a good idea to compare “shopping online” where the risk is hundreds of dollars, with “putting IP into the cloud” where the risk is thousands and millions of dollars.

    It is obvious that we will see lots of major data leaks from clouds within 3-5 years.

  • Granted, it all relative to who takes the risk. An individual actively using the web for her personal banking, investment, and shopping, as well as for her private interests (email, Facebook, etc), is exposing herself to risks ranging from unwanted spam to identity theft. For a company, the worst case is that its confidential data, which includes IPs, are compromised. Individuals gradually adopted web-based services for banking and other sensitive services. Companies gradually adopted cloud-based CRMs and other business operation services. I don’t see why companies will not come to use cloud-based services for more sensitive operations. The feel for security brought by having its own data center on its ground is somewhat misleading. It is as secured as the weakest link, which is people.

  • We see things less as private cloud vs public cloud, and more as creating ecosystems using the cloud as the connectivity. Some designers (and their organizations) will want to create their own private ecosystems to add their own value and to control their relationships with their customers in new ways. And some will use the obvious candidates mentioned in this thread.

    Our stance (at the risk of getting too commercial here) is to provide the tools and solutions so that the mainstream electronics designer can create his own ecosystem if he wishes to. In short: EDA needs to be about more than just the means of designing the next great piece of hardware, using the next great set of design features. It’s early days, and many designers (as hinted at in the thread) question the cloud. I think it’s far beyond that now, and as Harry says, “how much” is the real question (to which I’d add, “how easy/difficult?”).

  • Olivier, IaaS (as per AWS or an internal cloud) would imply that the software vendor would be managing, integrating, instantiating, monitoring, etc. the servers, storage, firewalls, and network themselves. With Xuropa this is not the case. We are at the Platform (PaaS) level of the Cloud Computing stack, in that we take care of everything related to the cloud, including customer on-boarding, security, etc. All our customers need do is install their software on the servers.

  • Cloud computing is regarded as the assinine idea ever conceived inside it. After spending decades and dollars to find out tips on how to keep information secure on a LAN, we’re designed to jump for joy on the prospect of putting our information on the WWW/WAN (within the full control of others) and feel secure?

  • The info travels through the WWW, but using secured channels –https, ssh, sftp, VPN. The data stored in the cloud can be encrypted, and decrypted data is kept in RAM only (it is never written on a disk but after encryption). There are many ways of setting up a client/server architecture so that:
    - breaching the https connection to snoop on the login session cannot reveal the customer password, or any of the encryption keys.
    - stealing the hard drive from the cloud provider cannot reveal anything because no data is stored in clear.

    Bottom line is, without the decryption keys, there is not much that can be done. Since no sensitive data (e.g., password, key) travel in clear between the customer’s browser and the cloud’s server, snooping cannot lead to any breach. The only way is to get the user password. More and more security breach are not due to some sophisticated attack on the portal or/and server, but are due to “social snooping”. Anybody can get a surprisingly rich set of information on people, using Facebook or any social sharing services. From there, once can try to determine a user password. Another technique is to send email with attachment that contains malware, which will release information or scan the keystroke of the user. The user is really the weakest link in that security chain.

  • There was immediate impression of ‘like’ reading on EDA on Cloud.

    It will be common and first level inclination to SAAS for EDA on web/cloud.

    Here is an attempt to next-step.

    HOTBENCH.NET

    A Web based utility service to speed up test-bench development in VHDL & Verilog. Engineers intuitively develop HDL test-benches with just click of mouse in substantially ‘Very Short Time’.

    URL: http://www.hotbench.net/

    -Amol

Leave a Reply

  

  

  

You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>