Standard cell libraries and embedded memory compilers - Added Value

 

Proven expertise

  • When Dolphin Integration delivered it’s first memory instance more than twenty five years ago, it was one of the first IP providers.
  • Today, Dolphin is one of the top 5 independent library providers.
  • Our engineering team currently includes about 150 engineers.
  • Our libraries benefit from proprietary techniques, often patented, for ensuring the best performance trade-offs.
  • The architecture of our libraries have proven their robustness in multiple foundries, IDMs, to major players and specialized foundries.

Ensuring SoC silicon success on first pass

  • Our libraries benefit from our proprietary qualification process VFP™ (Virtual Fab Process), which ensures that simulations are predictive of the robustness and of the yield. VFP™ is the efficient substitute to the inadequate Trial-and-Error qualification process on silicon.
  • We guarantees that libraries perform in accordance with Specifications, through our unique "Trust Commitment" clause.
  • We provides Engineering Support to help SoC Integrators build an efficient memory subsystem.
  • Our engineers have developed a unique validation method to anticipate yield drop due to the mismatch.
  • Our engineers have developed a calculation method of the OCV to be applied on its cell libraries.

Completeness of offering

FlexLib is the comparable architecture for methodical constructions of voltage and energy islets including:

  • Reduced Cell Stem Library
  • SpRAM and DpRAM
  • Metal programmable ROM
  • SpRfile, DpRFile
  • Regulators, integrated switches… for structuring an efficient power management network (upstream and downstream regulators)
  • Detectors: BIST, BISD, ECC for structuring an efficient test network

A set of innovations and patented solutions

  • Low power architectures specifically designed to enable robust low voltage operation, with characterizations taking into account physical phenomena linked to low voltage
  • Design optimization for minimizing leakage of RAMs, standard cells... (LLA)
  • Bit-cell mastery for reducing power consumption of memory planes (divided by 4) while improving speed and reducing area (XAM)
  • New approach for optimizing Performance, Power and Area of any logic block through the fusion of library stems with optimal adaptation at "logic path level"
  • Patented SESAME BIV (Battery Interface Voltage) designed with 3.3 V transistors for direct battery supply
  • Patented D Flip Flop: less sensitive to clock edges and guarantying the functionality at very low voltage
  • Patented High density Flip Flop: the spinner cell system
  • Patented Multi-bit ROM bit-cell (tROMet)
  • Patented Dual-bit ROM bit-cell (1/2T sROMet)

Find the optimal performance trade-offs!

Dolphin Integration has continuously enriched their offering of embedded Memories and Standard Cell Libraries so as to offer Off-the-Shelf products from 180 nm down to 40 nm endowed with differentiated optimizations to address a diversity of SoC requirements.

Typical gains
High Density - Low Power Panoply
Core VDD +/-10%
Power Consumption (uA) at Nominal Voltage *
Area Reduction *
180 nm G
1.8 V
Divided by up to 2
Up to 20%
130 nm G
1.2 V
Save up to 40%
Up to 30%
65 nm LP
1.2 V
Divided by up to 2
Up to 10%

 

Typical gains
Low Power - Dual Voltage Panoply
Core VDD +/-10%
Power Consumption (uA) at Nominal Voltage *
Power Consumption (uA) at Low voltage *
180 nm G
1.8 V
1.1 V
Divided by up to 3.5
Divided by up to 7
130 nm G
1.2 V
0.9 V
Save up to 30%
Divided by up to 2

 

Typical gains
Ultra Low Leakage Panoply
Core VDD +/-10%
Leakage *
180 nm G
1.8 V
Divided by up to 1,000
130 nm G
1.2 V
Divided by up to 1,000

* Comparison with free libraries