Last quarter, when we introduced the new AMD Opteron™ 6200 Series processor, we talked about the ability to support something called “Load Reduced DIMM” (LRDIMM) which was a new memory technology that was not previously available.  At the time, the voice of the customer was saying “what is LRDIMM and why should I care?”

While many will never exceed the current limits on memory capacity, there is a new class of applications, from Memcached to Hadoop and data analytics that demand large memory footprints in order to operate.

Load Reduced DIMMs help bring new levels of memory scalability to AMD Opteron 6200 Series processors with their support for huge memory footprints. By adding a buffer, LRDIMMs can multiply ranks, allowing for greater capacity of memory to be installed on a server.

As you may recall, our AMD Opteron 6100 Series processors supported up to 256GB of memory per socket, but with the new LRDIMMs, the capacity on the new AMD Opteron 6200 Series reaches an astounding 384GB per socket, giving a two socket server the ability to reach up to 768GB of total addressable system memory.

But who wants that much memory? Well, we’ve all heard the story of “who’d ever want more than 640K”, and before too long, we might be laughing at today’s memory capacities.  With complex data, complex media and complex environments, being able to do more in memory is key.  As a silicon guy, sometimes I hate to admit the truth, that memory can have a bigger impact on the scalability of systems than the CPU.  Of course, this is why we designed a system around four channels of DDR-3 memory.

With more memory you can handle more users, load larger data sets, manipulate more information and deploy more virtual machines.  It is not uncommon in the database world to load large tables directly into memory for rapid lookup.  More memory is a good thing, and LRDIMMs are like memory on steroids (without all of the nasty side effects.)

Today, Inphi is announcing that they have validated the new by Inphi iMB™ memory for the AMD Opteron 6000 Series platform. According to Inphi:

“The Inphi iMBGS02A passed several rigorous validation tests by AMD across process, voltage temperature and frequency.  Now shipping in volume and fully compliant with the memory-buffer specification of the JEDEC Solid State Technology Association, the devices have now been verified for compliance on the AMD 6200 Series platform, for devices’ interoperability.”

This is great news for the world of database, cloud, HPC and technical computing where large capacity memory is key ingredient of a high performance system.

This memory buffer allows for speeds of up to 1600MT/s, making them a great solution for not only maximizing memory capacity, but for also doing it at high speeds.

John Fruehe is the Director of Product Marketing for Server, Embedded and FireStream products at AMD. His postings are his own opinions and may not represent AMD’s positions, strategies or opinions. Links to third party sites, and references to third party trademarks, are provided for convenience and illustrative purposes only.  Unless explicitly stated, AMD is not responsible for the contents of such links, and no third party endorsement of AMD or any of its products is implied.