Japanese SBI Sumishin Net Bank, commissioned NRI, a leading research group in Japan, and together with Dragonfly conducted a test of the NEM blockchain technology. They were testing a version of NEM technology called Mijin.   (Mijin is a private fork of NEM that has some optimizations to make it a private chain). They were testing Mijin with 15 second blocks at over 2 million transactions per day of different kinds spread out over 2.5 million accounts trying to simulate the load a bank might be under.  The transaction load was handled fine.  Then they tried to systematically knock out nodes, and it was still fine.  Then they tried to tamper with transactions and fool the network, and still it was fine.

They concluded that NEM technology was capable of handling the back end of a banking platform and would ultimately bring savings costs and better stability to the banking platform, but they have no further plans to make the switch from current infrastructure to a blockchain like NEM at this time due to the price and logistics of making the switch.  

Here is a link to the article posted by SBI on a private financial news service.  To view the article you must be a member of the Nikkei.  http://itpro.nikkeibp.co.jp/l1/re_nft/?Mr2VlAOeZa2NvjvJlD0USokA8IVR4JxjhnIAVFWBvij6ECIiU612Xr6IjZX9Byl0ppiBi69Qtpve%0D%0ACd3NZLVBAfKl0VjJQg5Rryoe9t%2BEqSwOaVVxaRkm221GuFSsT8GVkA72F%2FJRZqs%3D

And here is a link to a different site (through Google translate) discussing the test and rehashing its results.  https://translate.google.com.au/translate?sl=auto&tl=en&js=y&prev=_t&hl=en&ie=UTF-8&u=http%3A%2F%2Fmijin.io%2Fja%2F588.html&edit-text=

And this test started in December and ran for a few months, but since December NEM/Mijin has been under development getting better but only the older version was tested.  

why does a cryptocurrency need approval from a bank?? The purpose of crypto is to compete with banks not help them.

Bitcoin and clones can do 2 million transactions per day using 8M blocks, this doesn't make them scalable.

@freshman777 and is there an article released by a third party that was testing Bitcoin and stressing it at those rates, not just seeing if it could get there in ideal conditions? Because for a blockchain to really make progress as a cryptocurrency and as a blockchain platform, it needs to scale. 

Yes, it needs to scale. You haven't proved NEM can scale by showing it can pump 2 million transactions per day. Scalability is much more than that.

Why would those blocks be orphaned?

What is required for scaleability then according to you?

NEM has a long line of firsts.  

• A new consensus mechanism called POI.
• First ever P2P time sync of nodes. (great for security of the chain)
• First use of the Eigentrust++ as a security clustering algorithm allowing node reputations to be given and shared to their neighbor nodes in clusters, making it very difficult for a node to give false information consistently to the network. (great for security of the network)
• First use of a new kind of transaction spam guard. It targets the specific accounts doing the spamming and raises their transaction fees, while letting the global transaction fee rate stay low. (good for the chain)
• First blockchain with a multisig 2.0 protocol. NEM’s version of multisig doesn’t make new accounts for a multisig address, instead it takes an existing account and makes a contract over it with other accounts. This form of multi-user accounts can make any form of M of N with ease and can also can have the M or the N edited with ease as it is simply just updating the meta data of the blockchain based contract. This form of multisig also has blockchain based notifications so transactions or updates of the M of N are pushed to the wallets as transactions to sign. (great security and ease of use for the user. simply the easiest and most flexible multisig in crypto)
• NEM’s multisig is so easy and flexible to use, it can be used as a foundational tool for DAO’s.
• Mosaics are the most advanced token creation feature in crypto allowing for many advanced 2.0 features.  When creating personalized blockchain based tokens, the mosaic can be created as mutable, immutable, divisible, not divisible, levies, transferable, not transferable, descriptions, and a name as a part of namespaces. (great for businesses)
• Along with mosaics, NEM will also released its new naming service, Namespace, that is also be the most advanced and most integrated naming service of any blockchain. (great for reputation of businesses that build on NEM as all projects can be under an umbrella name or or even sub-domains)

And there is more.....

Messaging, which can be done unencrypted/encrypted/hex. (great for users and programmers)
A light wallet that can easily be adapted by third party devs as it is pure pure JS/HTML/CSS. Any wallet may interact with any node on the network safely.  (great for users. wow!)
Node Rewards supporting Supernodes. (added network strength)
And........ EASY to use universal APIs for just about everything a person can do on NEM. This means it is a great platform for businesses to build their applications on.


Yes, as stated in the OP it was conducted on a private fork.  That said NEM is looking how to scale the right way. It is very important to us.  In fact everything that is built on NEM is built considering "will this scale?", and the recent test is one proof of it.  We look forward to in the future making more announcements about this and hope to add more "firsts" to the NEM list.

The test wanted to simulate realistic conditions, not a lab setting or intranet.  The nodes were geographically separated.  

NEM Technology was stress tested under bank conditions and received positive review. Nice. That's no small feat. Good job NEM devs and team. This definitely a testament not only for NEM's tech/team but all of crypto.

Yes. NEM has a team of committed devs working hard each day. There will be more improvements on the way to version 1.0.