The case for cold DRAM.
Cryogenic, superconducting digital processors offer the promise of greatly reduced operating power for server-class computing systems. This is due to the exceptionally low energy per operation of Single Flux Quantum circuits built from Josephson junction devices operating at the temperature of 4 Kelvin. Unfortunately, no suitable same-temperature memory technology yet exists to complement these SFQ logic technologies. Possible memory technologies are in the early stages of development but will take years to reach the cost per bit and capacity capabilities of current semiconductor memory. We discuss the pros and cons of four alternative memory architectures that could be coupled to SFQ-based processors. Our feasibility studies indicate that cold memories built from CMOS DRAM and operating at 77K can support superconducting processors at low cost-per-bit, and that they can do so today.
To read more, click here.
Interest in the open-source ISA marks a significant shift among chipmakers, but it will require continued industry support to be successful.
Slowdown due to impact on timing, and dependencies between power, thermal and timing that may not be caught by signoff tools.
Access to source code makes it attractive for custom applications, but gaps remain in the tool flow and in software.
SoC bandwidth, integration expand as data centers use more FPGAs for machine learning.
Problems won’t derail next-gen litho, but could limit use and affect schedules.
Interest in the open-source ISA marks a significant shift among chipmakers, but it will require continued industry support to be successful.
Using the Von Neumann architecture for artificial intelligence applications is inefficient. What will replace it?
Assessing the reliability of a device requires adding more physical factors into the analysis, many of which are interconnected in complex ways.
Slowdown due to impact on timing, and dependencies between power, thermal and timing that may not be caught by signoff tools.
Improvements in power, performance and area are much more difficult to achieve, but solutions are coming into focus.
Shift to cloud model has kept power consumption in check, but that benefit may have run its course.
Old solutions don’t necessarily work anymore, particularly at advanced nodes and ultra-low voltage.
The number of critical design metrics is expanding, but the industry still grapples with their implications.
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
Leave a Reply