News

posted on 25 February 2008 10:29

AMD's former chip architect, Fred Weber, responsible for its Opteron decision, has launched MetaRAM, his stealth start-up, claiming its technology quadruples server memory capacity.

It works for workstations too and can, he says, enable significant price drops. MetaRAM technology works with both Intel and AMD processor technologies.

In more detail, DDR2 MetaSDRAM is memory technology which, the company says, significantly increases server and workstation performance while dramatically decreasing the cost of high-performance systems. Using it, a 250GB, four-processor server with 16 cores starts at under $50,000, an up to a 90 percent reduction in system cost – all without any system
modifications.

MetaSDRAM is currently available in R-DIMMs from Hynix Semiconductor, Inc. and SMART Modular Technologies. Servers and workstations from Appro, Colfax International, Rackable Systems and Verari Systems are expected in the first quarter of 2008.

The company says MetaSDRAM is a drop-in solution that closes the gap between processor computing power, which doubles every 18 months – and DRAM capacity, which doubles only every 36 months. Until now, the industry has addressed this gap by adding higher capacity, but not readily available, and exponentially more expensive DRAM to each dual in-line memory module (DIMM) on the motherboard.

The MetaSDRAM chipset, which sits between the memory controller and the DRAM, solves the memory capacity problem cost effectively by enabling up to four times more mainstream DRAMs to be integrated into existing DIMMs without the need for any hardware or software changes. The chipset makes multiple DRAMs look like a larger capacity DRAM to the memory controller.

A single DDR2 MetaSDRAM DIMM can have 4GB or 8GB of memory but is a drop-in substitute for a normal DIMM.

The result is “stealth” high-capacity memory that circumvents the normal limitations set by the memory controller. This new technology has accelerated memory technology development by 2-4 years.

The MetaSDRAM chipset combines four separate 1Gb DDR2 SDRAMs into a single virtual 4Gb DDR2 SDRAM which acts exactly as a monolithic 4Gb DDR2 MetaSDRAM would.

The DDR2 MetaSDRAM chipset is optimized for low power and high performance. Its features include:

- nWakeOnUse power management which improves the power efficiency of the DRAMs, thus enabling two to four times the memory to fit into a typical system’s power delivery and cooling capabilities.
- Dynamic command scheduler that ensures that the MetaSDRAM is compatible with the JEDEC DDR2 protocol.
- Low latency circuit design and an innovative clocking scheme, which allow the MetaSDRAM-based DIMMs to fit into existing memory controller designs.
- Unique split-bus stacked DRAM design that enables flexible access of the multiple DRAMs in a stack.

Weber, CEO of MetaRAM, said: “I’ve spent my career focused on building balanced computer systems and providing compatible and evolutionary innovations. With the emergence of multi-core and multi-threaded 64 bit CPUs, I realized that the memory system is once again the biggest bottleneck in systems and so set out to address this problem. MetaRAM’s new MetaSDRAM does just that by bringing breakthrough main memory capacity to mainstream servers at unprecedented price points, without requiring any changes to existing CPUs, chipsets, motherboards, BIOS or software.”

This looks pretty solid and almost perfectly timed to capitalise on the spread of virtual servers where more memory means more virtual machines.

Bryan Wolf, managing director, Enterprise Platforms, Intel Capital, said: “The rapid adoption of Quad-Core Intel Xeon processors and platform virtualization, combined with the growth of data intensive applications, is driving demand for increased server memory capacity. MetaRAM’s technology presented an opportunity for Intel to participate as both an investor and a strategic technology collaborator to deliver a compatible solution that enhances system performance.”

MetaRAM has 35 employees and came alive with its initial January, 2006, funding. Its first chipset was released in November last year. Its products are designed for high performance rack-mount servers and workstations that run compute-intensive applications such as CAD/EDA simulations, database transaction processing (OLTP), business intelligence, digital content creation, and virtualization.

These and other heavy workload applications are the backbone of industries like aerospace, automotive, financial services, animation, oil and gas exploration, and semiconductor design and simulation.