In the early 1980s, the emergent computer industry incorporated mathematician and physicists John Von Neumann's distributed theorized compute model. Von Neumann's theories were way ahead of his time and were conceived long before the personal computing era became a reality. The Von Neumann model enabled the notion that many smaller computers could scale and produce higher computer power than a single centralized expensive computer (e.g., mainframe). As the Digital Age began, the personal computer not only became powerful but also grew in presence in homes and offices, bringing the usefulness of applications. Overtime, the personal computer (PC) out grew just being a desktop device and expanded into the data center and morphed into servers. Servers in the data center transformed into the client-server market and the well-known distributed compute model that John Von Neumann theorized forty-five years prior became reality.
For decades the PC, laptops and servers have been known to use RISC, PowerPC, ARM® and x86 architectures for processing power (CPU), limited memory (e.g., Random Access Memory RAM) and Hard Disk (HDA) devices for storage media. As the digital era continued to expand, the content computers created continued to get richer, larger in density and drove yearly innovation and upgrades in computer processing power (CPU), RAM capacities and hard drive densities. There continues to be several detriments to this approach; (1) not all components are gaining performance while gaining density [Moore's Law]; (2) the I/O interfaces of these elements are not the same speed, creating I/O bottlenecks [Kryder's Law].
A well-known upgrade technique in the computer industry has been to upgrade a computers memory (RAM) to get more performance out of a machine. Conversely, memory (RAM) capacities have been limited by several key factors, the CPU processor, nanometer density limitations of silicon, and power dissipation. By today's standards the largest memory module available is only 128 GB in capacity in contrast to the largest computer hard drive is 6 TB in capacity. In this example the hard drive is 93.75× larger than the memory module; this is the density issue. Contrariwise, the maximum input/output (I/O) transfer speed for memory modules (i.e., RAM) is currently 56.7 GB per sec, and the maximum I/O transfer speed for a Serial Attached SCSI (SAS-II) interface is currently 750 MB per sec. Thus, the memory module is 76.8× faster than today's SAS-II hard drive.
Under light computing loads, one might not notice this imbalance or battle of density vs. performance. However under a heavy computing load there is no equalizing this major imbalance of density vs. performance and I/O bottlenecks inevitably will occur. These eventually will slow the entire computing operation to the speed of the hard drive. The futile attempt to avoid this is to add more systems at the problem and rewrite applications to further distribute applications over more processor cores.
The answer to this quintessential problem would be to add more memory (RAM) and write application algorithms to alleviate the bottlenecks.
Nevertheless, the next challenge materializes, cost. Memory (RAM) in general can be very expensive depending of the density of the RAM module. A real world example of how expensive RAM is that the largest available memory module currently available is 64 GB. A single 64 GB RAM module currently sells for about $1,000.00 USD per module. The average x86 server motherboard currently sells for about $700.00 USD and can use up to 16 or 24 RAM modules. By fully populating an inexpensive x86 motherboard with 16 modules currently would cost about $16,000.00 USD; this makes RAM about 20 times more expensive than the inexpensive motherboard and would yield only 1 TB of RAM.
In an unflawed world, computers would need only memory (RAM) and high speed processors. If the challenge of density and cost did not exist, then computers without storage devices would be possible. The hurdle becomes how a memory modules (RAM) functions. All memory modules today are considered a volatile technology, meaning that when you power off a compute system, the memory losses, power and the memory becomes erased. Storage device media of today do not have this issue when the power is removed, storage device media retain the information that had been written to them. When you combine all of the factors of density, performance, cost and volatility, one can quickly deduce the reality of a computer with only CPU and RAM has been unachievable.
What is needed is an improved computing system to overcome the drawbacks the conventional art described above.
Embodiments in accordance with the present disclosure provide an inexpensive computer computing and storage apparatus that relies upon CPU and RAM, without a need for a magnetic storage device such as a conventional rotating hard drive. Embodiments provide a computing and storage apparatus and system that provides a quantum leap beyond the above mentioned obstacles surrounding computing and storage. Embodiments in accordance with the present disclosure enable a computer that may have a 57.6 GB constant I/O level that is 76.8× faster than any x86 and high performance computer in existence today.
Embodiments in accordance with the present disclosure provide a system and method to encode and decode raw data. The method to encode includes receiving a block of uncoded data, decomposing the block of encoded data into a plurality of data vectors, mapping each of the plurality of data vectors to a bit marker; and storing the bit marker in a memory to produce an encoded representation of the uncoded data. The method to decode includes retrieving a plurality of bit markers from a memory, mapping bit markers in the plurality of bit markers to respective data vectors, combining the respective data vectors with a block of uncoded data to produce a composite uncoded data block; and producing the uncoded composite data block as the decoded data.
The preceding is a simplified summary of embodiments of the disclosure to provide an understanding of some aspects of the disclosure. This summary is neither an extensive nor exhaustive overview of the disclosure and its various embodiments. It is intended neither to identify key or critical elements of the disclosure not to delineate the scope of the disclosure but to present selected concepts of the disclosure in a simplified form as an introduction to the more detailed description presented below. As will be appreciated, other embodiments of the disclosure are possible utilizing, alone or in combination, one or more of the features set forth above or described in detail below.
The above and still further features and advantages of the present invention will become apparent upon consideration of die following detailed description of embodiments thereof, especially when taken in conjunction with the accompanying drawings wherein like reference numerals in the various figures are utilized to designate like components, and wherein:
The headings used herein are for organizational purposes only and are not meant to be used to limit the scope of the description or the claims. As used throughout this application, the word “may” is used in a permissive sense (i.e., meaning having the potential to), rather than the mandatory sense (i.e., meaning must). Similarly, the words “include”, “including”, and “includes” mean including but not limited to. To facilitate understanding, like reference numerals have been used, where possible, to designate like elements common to the figures. Optional portions of the figures may be illustrated using dashed or dotted lines, unless the context of usage indicates otherwise.
In the following detailed description, numerous specific details are set forth in order to provide a thorough understanding of embodiments or other examples described herein. In some instances, well-known methods, procedures, components and circuits have not been described in detail, so as not to obscure the following description. Further, the examples disclosed are for exemplary purposes only and other examples may be employed in lieu of, or in combination with, the examples disclosed. It should also be noted the examples presented herein should not be construed as limiting of the scope of embodiments of the present invention, as other equally effective examples are possible and likely.
As used herein, the term “module” refers generally to a logical sequence or association of steps, processes or components. For example, a software module may comprise a set of associated routines or subroutines within a computer program. Alternatively, a module may comprise a substantially self-contained hardware device. A module may also comprise a logical set of processes irrespective of any software or hardware implementation.
A module that performs a function also may be referred to as being configured to perform the function, e.g., a data module that receives data also may be described as being configured to receive data. Configuration to perform a function may include, for example: providing and executing computer code that performs the function; providing provisionable configuration parameters that control, limit, enable or disable capabilities of the module (e.g., setting a flag, setting permissions, setting threshold levels used at decision points, etc.); providing a physical connection, such as a jumper to select an option, or to enable/disable an option; attaching a physical communication link; enabling a wireless communication link; energizing a circuit that performs the function (e.g., providing power to a transceiver circuit in order to receive data); and so forth.
System 100 includes a processor 102, which may be a general-purpose processor such as Xcon®, Intel Core i7®, i5®, i3®, or processors from Advanced Micro Devices® (AMD) such as Athlon64®, and the like. In other embodiments, processor 102 may be a graphics processing unit (GPU). In the functional block diagram of
Processor 102 may use several internal and external buses to interface with a variety of functional components. System 100 includes communication bus 105 that links processor 102 to memory controller 106. Memory controller 106 may also be referred to as a northbridge. Communication bus 105 may be implemented as one of a front side bus (FSB), a Non-uniform memory access (NUMA) bus, an EV6 bus, a Peripheral Component Interconnect (PCI) bus, and so forth.
System 100 further includes a nonvolatile memory 122 (e.g., a CMOS memory) coupled to processor 102. CMOS memory 122 may include a basic input/output system (BIOS) 124, which helps manage low-level communication among computer components, and may include storage of computer code to perform a power-on self-test. Conventionally, a power-on self-test may include a test of the data integrity of installed RAM.
Memory controller hub 106 typically handles communications between processor 102 and various high-speed functional components such as external RAM memory installed in dual, in-line memory module (DIMM) slots 108a, 108b via communication bus 107, and video graphics card 110 via communication bus 109. Communication buses 107 and 109 may be high-speed interfaces, such as Peripheral Component Interconnect Express (PCI) or Accelerated Graphics Port (AGP). Memory controller hub 106 may also handle communications between processor 102 and controller hub 114, via communication bits 112. Controller hub 114 may also be known by other names such as a southbridge, an I/O Controller Hub (ICH), a Fusion Controller Hub (FCH), a Platform Controller Hub (PCH), and so forth. Controller hub 114 in turn manages further communication with additional and/or slower I/O devices or interfaces such as USB ports 131, disk drives 132 with standard interfaces (e.g., ATA/SATA, mSATA, SAS, etc.), Ethernet transceivers 133, audio ports 134, other PCI devices 135, and so forth.
In some configurations of system 100 (not illustrated), processor 102 is designed to bypass memory controller 106 and communicate directly with controller hub 114 via a Direct Media Interface (DMI). Such configurations also may integrate the functions of processor 102 and memory controller 106 into a single IC 116. In such configurations, controller hub 114 is typically a Platform Controller Hub (PCH).
Although the memory chips that make up RAM memory installed in DIMM slots 108a, 108b may have a very high maximum access speed (e.g., about 57 GBytes/sec), communication bus 109 normally cannot support such fast speeds. For example, the speed of PCIe 4.0 in a 16-lane slot is limited to 31.508 GBytes/sec. AGP is slower still than PCIe. Therefore, communication bus 107 is a bottleneck that prevents faster memory access.
The bottleneck of memory access is one drawback of the conventional art. Other drawbacks described above of a conventional computer include the mismatch in storage size between the size of RAM memory (typically on the order of a few Gbytes) and the storage size of a conventional hard disk (typically on the order of a few Tbytes), and the relatively small storage size of RAM memory to the storage size of a conventional hard disk. Another drawback of the conventional art is the volatile nature of the RAM memory.
Embodiments in accordance with the present disclosure break the density issue that RAM has today. Embodiments in accordance with the present disclosure address these drawbacks of the conventional art by providing a novel hardware interface for storage units, and a novel driver interface for the hardware interface.
Outside of the CPU, RAM is the fastest element in x86 and x64 computing systems, so embodiments allows for the alignment of today's high speed RAM performance with a new method of gaining density. As this effect is applied, it completely changes the cost paradigm and allows low cost memory modules to replace the need for high-density, high cost memory modules.
A remaining issue is the volatility of standard memory modules. Since all RAM is volatile, it does not lend itself to becoming a long-term storage medium. Embodiments are similar to but differ from non-volatile RAM (NVRAM) technology, which circumvents the volatility issue found in standard DIMM devices.
Embodiments in accordance with the present disclosure use a basic inexpensive x64 motherboard that can be powered by Intel® or AMD® CPU processors. The motherboard has a modified CME and BIOS that gives it the intelligence required to be Non-Volatile Memory aware. In addition, the motherboard provides to each memory module a DC supply voltage (e.g., 1.2v, 1.35v, 1.5v, etc.) that may be used to charge environmentally-safe low-load, slow-drain capacitors. This design allows for shutdown state (e.g., loss of power or safe shutdown) to maintain data persistence within the memory module, thus making the memory module a viable long-term storage device.
System 200 further includes a nonvolatile memory 222 (e.g., a CMOS memory) coupled to processor 202. CMOS memory 222 may include a basic input/output system (BIOS) 224, which helps manage low-level communication among computer components, and may include storage of computer code to perform a power-on self-test. Conventionally, a power-on self-test may include a test of the data integrity of installed RAM. Embodiments in accordance with the present disclosure may include a modified power-on self-test (as compared to the power-on self-test of BIOS 124), such that the power-on self-test may skip the test for at least some predetermined memory modules, e.g., if the test would be incompatible with the nature of data stored in the predetermined memory module.
Embodiments in accordance with the present disclosure also address the RAM volatility shortcoming of the known art by coupling an energy source 219 with RAM storage unit 220. Energy source 219 may be incorporated with memory interface 218. Energy source 219 is a source of backup power, such that if an external power supply to RAM storage unit 220 is lost (e.g., by way of an AC power failure affecting the entire computing system 200, removal of a battery powering a mobile system 200, motherboard failure, etc.), energy source 219 may provide sufficient power in order to maintain integrity of data stored in RAM storage unit 220.
A destage process involves transferring data among physical media. Embodiments in accordance with the present disclosure implement a destage process by use of both hardware and software components. Hardware components include connector 208, memory interface 218, energy source 219 and RAM storage unit 220. Connector 208 may include a predetermined pin used to supply operating power to memory interface 218. Memory interface 218 may include limited processing power (e.g., a small CPU) to manage detection and notification processes such as for charging status of energy source 219, anomaly detection, and for LED lights (e.g., green, yellow, red). If a power loss is indicated, a switch may be activated to initiate a transfer to protected storage of data in memory interface 218 critical to system operation (e.g., configuration information, system state, etc.). Once destaging is complete, memory interface 218 may enter a reduced-power mode, and/or power off. Power for performing the destaging process may be supplied at least in part by energy source 219. Data integrity will be maintained by use of power from energy source 219.
If connector 208 is removed from its respective DIMM slot 108, connector 208 and/or memory interface 218 may include features to help ensure that if/when connector 208 is removed from a predetermined DIMM slot 108, that connector 208 when reinserted will be reinserted into the same respective DIMM slot 108. These features may act as security features, such that if incorrect reinsertion occurs, bits stored effectively will be scrambled and rendered unreadable.
Embodiments in accordance with the present disclosure also provide a restoration procedure. The restoration procedure is usable for recovery front a complete system failure or power loss. For example, if a motherboard fails, embodiments enter a low-power mode and repairs of the system (e.g., a motherboard replacement) or salvaging of components (e.g., removing of connector 208, memory interface 218, energy source 219 and RAM storage unit 220 for placement in a new unit). The restoration process includes re-installing memory modules into the same respective slot locations that they occupied in the defective unit. Memory modules may include physical or logical keying such that the memory modules will be unusable if installed in different slot locations within the replacement unit. For example, memory modules installed in different slot locations may produce scrambled bits (i.e., “bit scrambling”) if an attempt is made to read data from memory modules so inserted. The signature is associated with the module in the slot. So, for example, if you had a memory module originally installed in the B1 slot, and tried to reinstall it into the E1 slot, or vice versa, then the machine would not start. Data modules are matched to a slot number. A rationale for a hard association between memory modules and slot numbers is that attempting to restore data with memory modifies installed in the wrong slots may destroy data, so embodiments detect a wrong slot condition and prevent data access.
Backup and restoration procedures may be implemented by way of a module API (i.e., “chip calls”). API functions may include backup and restore. API functions may be triggered upon occurrence of certain predetermined event's, e.g., an “on-trigger” API call. For example, an application program may include a checkpoint, at which the system checks for an occurrence of a condition or an event that may pose a risk to the data. Upon such a detection, a backup API may be triggered, which may copy certain critical data, configuration information, metadata, etc., into a protected and nonvolatile memory such as a NAND memory. Examples of events that may trigger a backup include initiation of a boot process, recovery from a hardware or software fault, and so forth.
At system initialization, circuit power is supplied and the system components receive power. Energy source 219 will began to recharge. Status indicators (e.g., LEDs of various colors such as green, yellow, red, etc.) may be provided for the visual benefit of users. Status indicators may indicate progress at a first stage (e.g., performing a checksum). If a problem may be indicated, embodiments pay proceed to an alternate and/or more comprehensive procedure, e.g., checking the data image to carefully compare every bit and byte. Using conventional methods, this may be a very slow due to NAND speed limits, e.g., a ten-minute boot. Fault conditions may be detected and cause a system halt, module missing, module mismatch, etc. As noted earlier, if memory interface 218 and RAM storage unit 220 are not replaced correctly (e.g., wrong slot or wrong order), or are installed on a wrong system (i.e., one without required bitmarkers), data cannot be recovered. This behavior is needed in order to provide heavy security. A separate indicator may indicate when an API function is complete.
Voltage and power flow (e.g., 1.2v, 1.35v, 1.5v, etc.) is applied through connector 208 to a designated DIMM pin or alternate element, which in turn trader normal operation energizes energy source 219. Thermal design of memory interface 218 may include an element to cool energy source 219, e.g., by use of a heat sink.
Energy source 219 may have sufficient energy to maintain data integrity for a period of time of at least several months. Energy source 219 may be a large-capacitance capacitor known as a “super cap”, e.g., at least 600 Farads. Alternatively, energy source 219 may be a battery. However, a battery-based energy source such as a lithium battery is prone to catastrophic failure (e.g., arcing, fire) if damaged or in an internal short-circuit develops. Energy source 219 may be continuously charged under normal conditions by the main power to system 200 when energy source 219 is not supplying power to RAM storage unit 220.
System state may be restored, so long as system state is stored in RAM storage unit 220. In some embodiments, a separate backup energy source also may provide energy to other portions of a computing system (e.g., a processor and cache), such that a system state or other additional states may be preserved during a power outage. When the external power supply is restored, the computing system may be restarted or restored from a persistent or stateful state. In some embodiments, the system may cater or exist in a state of reduced power consumption while system state is preserved by the backup energy source.
Functions of memory interface 218 may further include monitoring a state of health of energy source 219, e.g., a voltage level since voltage levels may decay over time or in advance of a failure. Such a state of health may be communicated back to a monitoring system via communication bus 107. Memory interface 218 and RAM storage unit 220 may operate without requiring modification to cache memory 103.
System 200 further includes operating system 204, which is adapted to store and retrieve data to/from RAM storage unit 220. Operating system 204 includes data adaptation module 211 as part of a novel driver interface. Data adaptation module 211 executes bit generator software, which provides the functions of data adaptation module 211 described herein. The bit generation software may be loaded in real-time during the initialization process of processor 202. Conventional RAM memory (e.g., memory coupled to DIMM slot 108a) and/or cache memory 103 may be used to support functions of data adaptation module 211. When storing data, data adaptation module 211 adapts data to be stored in RAM storage unit 220 by encoding raw data into encoded data, and then storing the encoded data into RAM storage unit 220. Typically, for raw data of a predetermined size (i.e., a predetermined number of raw data bits), the encoded data is smaller, i.e., the encoded data may be represented by a smaller number of encoded data bits than the number of raw data bits. Data adaptation module 211 may store into RAM storage unit 220 an amount of data that, if represented in its raw form, would exceed the storage capacity of RAM storage unit 220. An effective storage capacity of RAM storage unit 220, e.g., as measured by an equivalent number of raw data bits, may exceed a physical storage capacity of RAM storage unit 220. Processor 202 may utilize conventional RAM memory and/or cache memory 103 in order to support operation of processor 202 for conventional functions, e.g., as a server.
The transfer of encoded data by data adaptation module 211 into RAM storage unit 220 may take place at a transfer rate that, if represented by the transfer of the equivalent raw data bits, would exceed the maximum data transfer rate of the processor, RAM storage unit 220, and the communication bus linking the processor to RAM storage unit 220.
Similarly, when retrieving stored data front RAM storage unit 220, data adaptation module 211 adapts data read from RAM storage unit 220 by decoding encoded data into raw data, and then providing the raw data for other uses, e.g., by processor 202 or application programs running on processor 202.
Data adaptation module 211 adapts the data at a sufficiently high speed such that resources spent encoding data (e.g., processing time by processor 202) is less than the resources saved (e.g., transmission time on communication bus 107) by transmission of encoded data rather than raw data to RAM storage unit 220. Similarly, the processing time spent decoding data is less than the time saving resulting front transmission of encoded data rather than raw data from RAM storage unit 220.
Embodiments in accordance with the present disclosure, when in production, may run a specialized thin OS in data adaptation module 211 that enables the platform to be a hybrid compute and storage system. The OS will also offer an embedded server virtualization platform to allow several virtual machines to run simultaneously on the platform. One of many examples of these would be a single system running ten to fifteen or more Microsoft Windows instances independently and simultaneously, all without over experiencing an I/O lag that otherwise would be inherent to conventional known virtual compute platforms.
Embodiments in accordance with the present disclosure may also be used when running extreme high-performance transaction processing found in complex database systems. Such applications enable the possibility of running a large portion of, or the entirety of, the system database purely in RAM.
Preliminary testing and/or simulation of embodiments indicate that a properly configured system could achieve upwards of 4 TB of extreme high speed RAM storage with as little 48 GB of NVRAM.
Processes to encode raw data, and to decode processed data to produce raw data, as described herein may be performed by invoking modules within data adaptation module 211. These modules may be invoked by operating system 204, or another application program executing on processor 202.
One embodiment of encoding of raw data is described in U.S. Patent Application Publication No. 2014/0223118 to Ignomirello (“the '118 Publication”), which is commonly assigned with the present application, and the entire content of which is hereby incorporated by reference.
Other embodiments may encode raw data by use of a Bit marker method, such as described in the '118 Publication, which enables the effective storage capacity of RAM storage unit 220 to become more elastic, and allowing the effective storage capacity and density to grow very quickly. Storage capacity may be elastic in the sense that an equivalent amount of raw data that can be stored is not a fixed value, but may change depending upon characteristics of the raw data, e.g., how well the raw data matches to the Bit markers. The raw data capacity is not controlled or known in advance by the operating system or by lower-level software such as the basic input/output system (BIOS). Embodiments may reduce the need for RAM chips of very high physical capacity, and concomitant very dense nanometer feature design, thus allowing for usage of lower-density and/or older generations of RAM chips for RAM storage unit 220.
A Bit marker may be derived from, or based at least in part from, a characteristic of the raw data, or a characteristic derived front the raw data. The characteristic of the raw data may be, for example, the type of raw data (e.g., an audio file, a graphics file, a video file, a ZIP file, a Word file, a PDF, etc.), a file format (e.g., for graphics files, bitmapped, GIF, TIF, JPEG, etc.); content of the file (e.g., for an MP3 audio file, whether it is an MP3 of classical music, jazz, rap, rock music, spoken words such as an audio book, etc.), attributes of the file (e.g., for an MP3 audio file, the bit rate, mono or stereo, metadata, total length, encoder used, dynamic range, time length, etc.), statistics of the file, (e.g., size, age, date modified, probability distribution of bytes or sequences of bytes within the file, etc.), and so forth. For example, an MP3 file may contain certain raw data bits (i.e., sequences of bytes) more often than a different type of file (e.g., a JPG file), therefore knowledge that a file is an MP3 file may affect the Bit markers selected and the raw data bits that each Bit marker represents. Data adaptation module 211 may adaptively change, over time or over amount of raw data processed, the Bit markers used to represent the raw data in response to changes in characteristics of the raw data, e.g., if the characteristics of the raw data become better known or are not stationary in a statistical sense.
A Bit marker may represent a vector of raw data, i.e., a predetermined sequence of raw data bits. The vector of raw data also may be referred to as a data segment. Different Bit markers may represent vectors of different length. Raw data may be decomposed into vectors of raw data and the vectors then mapped to a Bit marker. In some embodiments, concatenated vectors may represent the raw data. In other embodiments, vectors may overlap such that a combination of vectors (e.g., Boolean AND, OR, XOR, NAND, etc.) may represent the raw data. The raw data may be represented by a plurality of layers (e.g., for a graphics file, separate layers for each color component such: as R, G, B; or C, Y, M, K; or other logical layers, etc.), and Bit markers may be used to represent separately raw data bits within each color layer. In some embodiments, the raw data may be represented as a multidimensional data structure (e.g., a 2-D array, a 3-D cube, an N-dimensional structure, etc.), and a vector may represent a contiguous position of the multidimensional data structure (e.g., a rectangular portion of a 2-D array of raw data). A bit marker may be viewed as a translational function that translates between a vector pointer and the law vector data itself.
In some embodiments, knowledge of a Bit marker (e.g., the characteristic derived from the raw data), may be sufficient to generate model vectors to match to the raw data. For example, suppose that a set of raw bits (e.g., a file or information received from a communication link) represent video streaming media. It may be surmised, deduced, or otherwise configured that the raw bits represent video streaming media having particular characteristics (e.g., frame rate, resolution, pixel dimensions, color palette, etc.), and a Bit marker may be selected to indicate that the data is video streaming media of those characteristics. Knowledge that the Bit Marker represents video streaming media of those characteristics may be used to generate model vectors predictively matched to the characteristics of video streaming media, e.g., frame rate, resolution, pixel dimensions, color palette, etc. A benefit of such an embodiment is that once the Bit marker is known, encoding can be accomplished on the fly, e.g., by processing streaming media in real time. For the processing a file (e.g., an MP3 file, a DVD video disk, etc.), the processing may be accomplished without needing to read in the entire file (e.g., on a section-by-section basis), and needing only a buffer of a relatively modest size. Encoded data may be stored to RAM storage unit 220 as it is produced.
In some embodiments, an object may be a compound object, i.e., an object of one file type and yet encompass another file type. For example, an email object may include within it an embedded object such as an attached PDF, an attached ZIP file, an attached Word document, another email object, etc. Knowledge that an object is a compound object may be useful to help select vectors and Bit Markers appropriate for the embedded typo of object, and for separate portions of the compound object.
In some embodiments, raw data may be converted from a one-dimensional structure to a multidimensional structure by analyzing the raw data to determine useful boundaries between various sections prior to partitioning. The various sections after partitioning form the multidimensional structure. In some embodiments, the analysis may take into account available processor cache size, e.g., in order to help ensure that the analysis can be performed quickly.
In some embodiments, a Bit marker may represent a nucleus (i.e., a relatively small set of seed raw data bits, a starting value, a starting pattern, etc.) together with a replication rule for how to produce additional data from the seed, e.g., to expand the nucleus to an arbitrarily large number of raw data bits. The rule for producing additional data may include, e.g., replication a predetermined number of times or to a predetermined length, a fractal-based replication, run length replication, and so forth. Such a representation may be useful for a fractal-based mapping of raw data bits to a fractal pattern rendered at a particular size, for example, if a portion of a 2-D multidimensional raw data structure could be represented by a fractal pattern. As used by embodiments, fractal patterns for data encoding and decoding may exist for raw data in other dimensionality, e.g., 1-D (a linear pattern), 3-D, and so forth.
In some embodiments, a Bit marker may represent one or more DNA submarkers, together with a rule for combining the DNA submarkers. For example, a Bit marker may represent a vector V4 concatenated with the XOR product of V5 and V1. The resulting Bit marker then may act as a submarker for combining with other submarkers to form another marker.
Initially, vectors and their corresponding Bitmarkers may be stored in a vector field. The vector field is normally stored in a separate memory, apart from RAM storage unit 220. For example, if RAM storage unit 220 is coupled to DIMM socket 108b, then the vector field may be stored in a conventional DIMM memory coupled to DIMM socket 108a, or in cache memory 103, and so forth. A group of raw data bits may be transmitted to data adaptation module 211, which then maps or correlates the raw data bits or a portion thereof to vectors in the vector field. Corresponding Bit markers then may be retrieved from the vector field and stored in RAM storage unit 220 in place of the equivalent raw data bits. Bitmarkers may be reused or “amplified”, such that if the raw data includes multiple instances of raw data bits, the corresponding Bit marker may be stored in RAM storage unit 220 with an indication that it is used for the multiple instances of raw data bytes. Amplification refers to a ratio between the size of a bit marker (e.g., as measured by the number of bits or bytes) and the size of the equivalent raw data bits (or bytes) replaced by instances of the usage of the bitmarker.
In some embodiments, RAM storage unit 220 may be logically partitioned, such that one portion of RAM storage unit 220 (e.g., one physically addressable portion) may store Bitmarkers, while another portion of RAM storage unit 220 may operate a traditional memory. In some embodiments, RAM storage unit 220 coupled to one of the DIMM slots (e.g., DIMM slot 108b) may operate to store Bit markers, and a RAM module coupled to another DIMM slot (e.g.; DIMM slot 108a) may operate as conventional memory.
Other embodiments in accordance with the present disclosure include a bit generator which encodes long form data into short form bit Markers during data population and decodes short form bit Markers into long form data on the fly when requested.
In some embodiments, one or more patterns may be discerned in a set of raw data. The pattern may be, for example, a periodicity in the raw bits when the raw bits are expressed as a linear series of “0” and “1”. Patterns may also be in the form of a periodicity in the raw data when the raw data is expressed as a linear series of bytes, or a multidimensional periodicity when the raw data is expressed as a multidimensional set of raw data. The periodicity may be expressed as a frequency of a predetermined pattern in the raw data. Characteristics of the periodicity may include a frequency and a phase. Multidimensional data may be characterized independently in each dimension. A period (or conversely a frequency) may be expressed as raw bits per repeating cycle of the predetermined pattern, e.g., 256 bits per cycle, 65,536 bits per cycle, etc. A phase may indicate a starting point of the pattern with respect to a reference copy of the predetermined pattern (e.g., a dictionary copy). Embodiments may use the frequency and phase characteristics as identifying characteristics (e.g., a Bit marker, a fingerprint, etc.).
In some embodiments, a periodicity in a one-dimensional raw data may be modeled or analyzed as a multidimensional raw data. For example, a data trace representing an electrocardiogram includes an inherent periodicity represented by the heartbeat rate. The periodicity is not perfect, because the frequency (e.g., beats per minute) may change over time, and the exact shape of the electrocardiogram may change from one beat to another. Nevertheless, the electrocardiogram trace may be modeled as a multidimensional structure, with the electrocardiogram for one beat (e.g., a first beat, or a reference beat, or an ideal beat, etc.) representing one plane (i.e., two axes) of voltage versus time, and a third dimension representing ordinal beat number. Data may be encoded in part by analyzing the differences in the third dimension, after accounting for changes in factors such as frequency noted above.
In some embodiments, the pattern may represent a linear combination of one or mote basis functions or basis vectors. In mathematics, a basis function is an element of a particular basis for a function space. Every continuous function in the function space can be represented as a linear combination of basis functions. Similarly, every vector in a vector space can be represented as a linear combination of basis vectors. Basis vectors are known as a set of linearly independent vectors in a vector space, and every other vector in the vector space is linearly dependent on these vectors.
For example, the raw data may be decomposed into a combination of basis vectors. Each basis vector is a measurable binary pattern. Preferably, a basis vector should be very long compared to a Bit marker (i.e., a pointer to a basis vector in a vector map) used to reference the basis vector, but the basis vector may be shorter than the entire raw data. The representation of raw data as basis vectors may include one or more data pairs of (a) a Bit marker for a basis vector and (b) a starting position in the raw data of the basis vector. A gene pool is related to the vector map, in that the gene pool may include information regarding how to reconstruct raw data from vectors. For example, a gene pool may indicate a type of file to be reconstructed (e.g., an MP3 file), knowledge of which would be useful in reconstructing the file. Analogizing to a jigsaw puzzle, vectors may represent individual pieces of a jigsaw puzzle, and a gene pool may represent a photo that the entire jigsaw puzzle should look like when completed.
In some embodiments, raw data not represented by a basis vector may be deemed to be a predetermined value, e.g., a 0x00 or 0xFF byte. This may be useful if the raw data has a large number of consecutive bits or bytes of the predetermined value, e.g., a large number of consecutive 0x00 bytes.
In some embodiments, raw data may be decomposed into basis vectors that may at least partially overlap. Overlapping basis vectors may be combined by default as a Boolean OR, but other Boolean functions may be used (e.g., AND, XOR, NAND, etc.).
In some embodiments, the vector dictionary may be adaptive to changes in statistics of the raw data. For example, if the type of information being stored changes (e.g., from video to MP3), the statistics of the raw data may also change. Such changes may mean that certain basis vectors are used less often, and other basis vectors may be used more often. Embodiments may recognize changes in usage and update a basis dictionary appropriately, e.g., by culling some basis vectors and adding other basis vectors. The decoder will be aware of changes in the vector dictionary, e.g., by inclusion of a dictionary update.
In some embodiments, entries in the vector dictionary may have different, but fixed lengths. The lengths may depend upon statistics of the raw data. For example, Bit markers may have different lengths, such that Bit markers corresponding to more commonly-occurring vectors in the raw data may be shorter (i.e., the Bit marker comprises fewer bits) than Bit markers corresponding to less commonly-occurring vectors in the raw data. Alternatively, Bit markers may have equal lengths, but some Bit markers may correspond to a longer but more commonly-occurring vector of raw data than other Bit markers that represent shorter but less commonly-occurring vector of raw data.
In some embodiments, Bit markers may be represented in a tree and leaf paradigm, which may be inherently hierarchical. In this paradigm, each Bitmarker is represented by a leaf, with a size and/or position of the leaf in the tree corresponding to a characteristic of the Bit marker it represents. For example, a more commonly-occurring Bit marker may be represented as a smaller leaf or a leaf closer to the root of the tree. Conversely, a less commonly-occurring Bit marker may be represented as a larger leaf or a leaf farther from the root of the tree. The goal may be use leaves that are as small as possible, or to use leaves as close to the root as possible, or to use leaves that tend to minimize a mathematical function, such as a product of the size of the leaf times the number of times that the leaf is used.
Decoding data involves reading encoded data from RAM storage unit 220, and then performing functions to reverse the encoding processes. Decoding functions may be provided by modules within data adaptation module 211. For example, to restore data, a block of encoded data may be read from RAM storage unit 220. The block of encoded data may be temporarily stored in a high speed memory while decoding processes are performed. Decoding processes are provided by modules within data adaptation module 211. These modules may be called by operating system 204.
More particularly, when retrieving data from RAM storage unit 220 and decoding it, data adaptation module 211 adapts data to be stored in RAM storage unit 220 by decoding raw data into decoded data. Conventional RAM memory (e.g., memory coupled to DIMM slot 108a) and/or cache memory 103 may be used to support decoding functions of data adaptation module 211.
Process 300 begins at step 301, at which a block of raw data to be stored is received from an application program that intends to store the raw data. The raw data may be in the form of a file, a streaming media, a fixed-size or variable-size block of data, and so forth.
Next, process 300 transitions to step 303, at which portions of the raw data received in step 301 may be mapped or matched to candidate vectors of raw data. The candidate vectors may be stored as a table of (marker, vector) pubs in conventional memory. The goal is to represent each bit or byte in the raw data by at least one vector. Certain raw data bytes such as 0x00 or 0xFF may be deemed to be a default value, and for any raw data bytes equal to the default value, it is optional to represent the default bytes with a vector.
Next, process 300 transitions to step 305, at which vectors determined in step 303 may be mapped to a respective bit marker from the table of (marker, vector) pairs. The bit marker is a short way to refer to the associated vector.
Next, process 300 transitions to step 307, at width the bit marker from the table of (marker, vector) pairs is stored in memory, such as RAM storage unit 220.
Process 350 begins at step 351, at which a block of encoded data to be decoded is read from a memory, such as RAM storage unit 220. Addresses may be managed by virtual address adjustment methods and tables, as known to persons of skill in the art.
Next, process 350 transitions to step 353, at which bit markers are extracted from the encoded data.
Next, process 350 transitions to step 355, at which the extracted bit markers from step 353 are searched for in the table of (marker, vector) pairs.
Next, process 350 transitions to step 357, at which a raw data vector is extracted from an entry in the table of (marker, vector) pairs, corresponding to the extracted bit marker from step 353.
Next, process 350 transitions to step 359, at which the extracted raw data vectors front step 357 are combined to form reconstructed decoded data. If the combined raw data vectors do not cover all addresses within an entire expected address range of the reconstructed decoded data, the uncovered addresses may be deemed to take on a default value in the decoded data, e.g., 0x00 or 0×FF bytes.
When analyzing the I/O capability, conventional systems may allow for continuous I/O speeds up to 57.6 GB per second. In contrast, for a system in accordance with an embodiment of the present disclosure, the system tested with Intel Ivy Bridge 2697 v2 processors, embodiments may have 24 physical process cores and up to 40 hyper threaded cores, 6144 KB of 1.2 processor cache, 60 MB of L3 processor cache all at 5.4 GHz with a boost capability of 76 GHz. Comparing performance of the embodiments vs. any other Ivy Bridge 2697 v2 server shows an I/O increase of 76.8× faster. Commensurate performance gains are achievable with other computing environments, including Haswell motherboard architectures and DDR4 memory.
While the foregoing is directed to embodiments of the present invention, other and further embodiments of the present invention may be devised without departing from the basic scope thereof. It is understood that various embodiments described herein may be utilized in combination with any other embodiment described, without departing from the scope contained herein. Further, the foregoing description is not intended to be exhaustive or to limit the invention to the precise form disclosed. Modifications and variations are possible in light of the above teachings or may be acquired from practice of the invention. Certain exemplary embodiments may be identified by use of an open-ended list that includes wording to indicate that the list items are representative of the embodiments and that the list is not intended to represent a closed list exclusive of further embodiments. Such wording may include “e.g.,” “etc.,” “such as,” “for example,” “and so forth,” “and the like,” etc., and other wording as will be apparent from the surrounding context.
No element, act, or instruction used in the description of the present application should be constructed as critical or essential to the invention unless explicitly described as such. Also, as used herein, the article “a” is intended to include one or more items. Where only one item is intended, the term “one” or similar language is used. Further, the terms “any of” followed by a listing of a plurality of items and/or a plurality of categories of items, as used herein, are intended to include “any of,” “any combination of,” “any multiple of,” and/or “any combination of multiples of” the items and/or the categories of items, individually or in conjunction with other items and/or other categories of items.
Moreover, the claims should not be read as limited to the described order or elements unless stated to that effect. In addition, use of the term “means” in any claim is intended to invoke 35 U.S.C. § 112, ¶6, and any claim without the word “means” is not so intended.
This application is a continuation of U.S. patent application Ser. No. 15/783,838, filed on Oct. 13, 2017, which is a continuation of PCT Patent Application No. PCT/US2016/025988, filed on Apr. 5, 2016, which is a continuation of U.S. patent application Ser. No. 14/804,175, filed on Jul. 20, 2015, which claims the benefit of U.S. Provisional Application Ser. No. 62/148,160, filed on Apr. 15, 2015. Each of the foregoing applications is hereby incorporated herein by reference in its entirety.
Number | Name | Date | Kind |
---|---|---|---|
3016527 | Gilbert et al. | Jan 1962 | A |
3739410 | Fortin | Jun 1973 | A |
3926346 | Dub et al. | Dec 1975 | A |
4183464 | Babb | Jan 1980 | A |
4286256 | Langdon, Jr. et al. | Aug 1981 | A |
D264844 | Tuck | Jun 1982 | S |
4420767 | Hodge et al. | Dec 1983 | A |
4560977 | Murakami et al. | Dec 1985 | A |
4625710 | Harada et al. | Dec 1986 | A |
4635141 | Coulter | Jan 1987 | A |
D304193 | Lotz | Oct 1989 | S |
D308671 | Aderman et al. | Jun 1990 | S |
4931340 | Baba et al. | Jun 1990 | A |
D319765 | McCord | Sep 1991 | S |
5168171 | Tracewell | Dec 1992 | A |
D341866 | Ross | Nov 1993 | S |
5300930 | Burger et al. | Apr 1994 | A |
D353182 | Rubin | Dec 1994 | S |
D357461 | Zaplatynsky et al. | Apr 1995 | S |
D362243 | Spechts et al. | Sep 1995 | S |
D366246 | Simek et al. | Jan 1996 | S |
D372909 | Naufel | Aug 1996 | S |
D377482 | Geck | Jan 1997 | S |
D382880 | Cienkus et al. | Aug 1997 | S |
D391938 | D'Agaro | Mar 1998 | S |
D393455 | D'Agaro et al. | Apr 1998 | S |
5748921 | Lambrecht et al. | May 1998 | A |
5751549 | Eberhardt et al. | May 1998 | A |
5784611 | Thantrakul | Jul 1998 | A |
5785799 | Culnane et al. | Jul 1998 | A |
D396690 | Fryers | Aug 1998 | S |
D397097 | McNally | Aug 1998 | S |
5797143 | Buxton | Aug 1998 | A |
5818877 | Tsai et al. | Oct 1998 | A |
5867180 | Katayama et al. | Feb 1999 | A |
5872530 | Domyo et al. | Feb 1999 | A |
D407079 | Wu | Mar 1999 | S |
5878240 | Tomko | Mar 1999 | A |
5890794 | Abtahi et al. | Apr 1999 | A |
D409171 | Wu | May 1999 | S |
5923860 | Olarig | Jul 1999 | A |
D415747 | Wu | Oct 1999 | S |
D421428 | Wu | Mar 2000 | S |
D422993 | Storti et al. | Apr 2000 | S |
D423479 | Alo et al. | Apr 2000 | S |
6052785 | Lin et al. | Apr 2000 | A |
6067227 | Katsui et al. | May 2000 | A |
D431241 | Armstrong et al. | Sep 2000 | S |
6182601 | Baatz | Feb 2001 | B1 |
D439566 | Ohanian | Mar 2001 | S |
6243831 | Mustafa et al. | Jun 2001 | B1 |
6275173 | Wu | Aug 2001 | B1 |
D447484 | Reznikov et al. | Sep 2001 | S |
6297742 | Canada et al. | Oct 2001 | B1 |
6297753 | Hayami | Oct 2001 | B1 |
6310564 | Fujimoto | Oct 2001 | B1 |
6326781 | Kunde et al. | Dec 2001 | B1 |
6373695 | Cheng | Apr 2002 | B1 |
6396027 | Wu | May 2002 | B1 |
6401807 | Wyler et al. | Jun 2002 | B1 |
D462075 | Hillyard et al. | Aug 2002 | S |
6430351 | Takahashi et al. | Aug 2002 | B1 |
6442031 | Liu | Aug 2002 | B1 |
6465969 | Murasko et al. | Oct 2002 | B1 |
6483439 | Vukosic | Nov 2002 | B1 |
6494780 | Norbury, Jr. | Dec 2002 | B1 |
D471912 | Camarota et al. | Mar 2003 | S |
6541921 | Luciano, Jr. et al. | Apr 2003 | B1 |
D475061 | Huang | May 2003 | S |
6560599 | Boa et al. | May 2003 | B1 |
6605778 | Dorfler et al. | Aug 2003 | B2 |
6614369 | Widmer | Sep 2003 | B1 |
6621935 | Xin et al. | Sep 2003 | B1 |
6626557 | Taylor | Sep 2003 | B1 |
6644481 | Dean et al. | Nov 2003 | B2 |
D483374 | Hung | Dec 2003 | S |
D483378 | Mundt | Dec 2003 | S |
6658435 | McCall | Dec 2003 | B1 |
6674908 | Aronov | Jan 2004 | B1 |
6683545 | Strasser | Jan 2004 | B2 |
D490083 | Wu | May 2004 | S |
6737299 | Jiang | May 2004 | B1 |
6785133 | Barringer et al. | Aug 2004 | B2 |
6785815 | Serret-Avila et al. | Aug 2004 | B1 |
6829695 | Ross | Dec 2004 | B1 |
6844824 | Vukosic | Jan 2005 | B2 |
D503717 | Kawamura | Apr 2005 | S |
6904484 | Nelson | Jun 2005 | B1 |
6905228 | Takeyasu et al. | Jun 2005 | B1 |
6926972 | Jakobi et al. | Aug 2005 | B2 |
6956745 | Kerrigan et al. | Oct 2005 | B2 |
D513751 | Alo et al. | Jan 2006 | S |
6999319 | Wu et al. | Feb 2006 | B2 |
7030837 | Vong et al. | Apr 2006 | B1 |
7064265 | Cochrane | Jun 2006 | B2 |
7111961 | Trenchard et al. | Sep 2006 | B2 |
D529915 | Scicluna et al. | Oct 2006 | S |
D530052 | Berti | Oct 2006 | S |
D530343 | Maruyama et al. | Oct 2006 | S |
D530704 | Cortes | Oct 2006 | S |
7145776 | King, Jr. et al. | Dec 2006 | B2 |
D535298 | Cochrane | Jan 2007 | S |
7164581 | Carullo et al. | Jan 2007 | B2 |
7174436 | Langendorf et al. | Feb 2007 | B1 |
7231560 | Lai et al. | Jun 2007 | B2 |
7246037 | Evans | Jul 2007 | B2 |
7290116 | Grohoski et al. | Oct 2007 | B1 |
D555159 | Cox et al. | Nov 2007 | S |
D556204 | Tosh et al. | Nov 2007 | S |
7298535 | Kuutti | Nov 2007 | B2 |
D565573 | Alo | Apr 2008 | S |
7356679 | Le et al. | Apr 2008 | B1 |
7357530 | Wang et al. | Apr 2008 | B2 |
7363331 | Karasudani et al. | Apr 2008 | B2 |
D570851 | Gundogan et al. | Jun 2008 | S |
7371975 | Oettinger et al. | Jun 2008 | B2 |
D579455 | Chu | Oct 2008 | S |
7463163 | Wu | Dec 2008 | B2 |
7485510 | Zhu et al. | Feb 2009 | B2 |
7492607 | Nguyen et al. | Feb 2009 | B2 |
D590831 | Evans | Apr 2009 | S |
7515413 | Curtis | Apr 2009 | B1 |
7516272 | Felton et al. | Apr 2009 | B2 |
7539026 | Finnerty et al. | May 2009 | B2 |
D593565 | Dearborn et al. | Jun 2009 | S |
7587408 | Snapp et al. | Sep 2009 | B2 |
7620784 | Panabaker | Nov 2009 | B2 |
D609236 | Matsumoto et al. | Feb 2010 | S |
7726837 | Ewert | Jun 2010 | B2 |
7731180 | Ohnishi | Jun 2010 | B2 |
D619108 | Lin | Jul 2010 | S |
7794124 | Hulsey et al. | Sep 2010 | B2 |
7822946 | Sharma | Oct 2010 | B2 |
D631480 | Vaughan | Jan 2011 | S |
D631630 | Berti | Jan 2011 | S |
7921088 | Mittal | Apr 2011 | B1 |
7925389 | Bradicich et al. | Apr 2011 | B2 |
7930473 | Rajan | Apr 2011 | B2 |
D643434 | Grady, IV et al. | Aug 2011 | S |
7991941 | Fortin | Aug 2011 | B2 |
7996361 | Shah et al. | Aug 2011 | B1 |
8001079 | Lu et al. | Aug 2011 | B2 |
8006108 | Brey et al. | Aug 2011 | B2 |
8009069 | Chen et al. | Aug 2011 | B2 |
D646954 | Azelton et al. | Oct 2011 | S |
8078789 | Hinrichs | Dec 2011 | B2 |
8117028 | Murashima | Feb 2012 | B2 |
D656250 | Forster et al. | Mar 2012 | S |
D656502 | Cacioppo et al. | Mar 2012 | S |
8129742 | Lin et al. | Mar 2012 | B2 |
8139945 | Amir et al. | Mar 2012 | B1 |
8164490 | Itani | Apr 2012 | B2 |
D659136 | Matsumoto | May 2012 | S |
D659137 | Matsumoto | May 2012 | S |
D659653 | Lin | May 2012 | S |
D661696 | Takada | Jun 2012 | S |
D662501 | Lohman | Jun 2012 | S |
D662504 | Lohman | Jun 2012 | S |
8200923 | Healey et al. | Jun 2012 | B1 |
D663299 | Corke et al. | Jul 2012 | S |
8214425 | Provencal | Jul 2012 | B2 |
8229898 | Pype | Jul 2012 | B1 |
8237594 | Suzuki et al. | Aug 2012 | B2 |
8238091 | Chang et al. | Aug 2012 | B2 |
8238840 | Lio et al. | Aug 2012 | B2 |
8265323 | Stiehl et al. | Sep 2012 | B2 |
8281013 | Mundkur et al. | Oct 2012 | B2 |
8290160 | Steeger | Oct 2012 | B1 |
8310830 | Zheng et al. | Nov 2012 | B2 |
8314555 | Huang et al. | Nov 2012 | B2 |
8322040 | Jimenez | Dec 2012 | B2 |
8334838 | Sadahiro et al. | Dec 2012 | B2 |
D676039 | Wallace et al. | Feb 2013 | S |
D676852 | Lohman | Feb 2013 | S |
8366464 | Figuerado et al. | Feb 2013 | B1 |
D678273 | Kuehn et al. | Mar 2013 | S |
D681635 | Wang | May 2013 | S |
D682267 | Mase | May 2013 | S |
8458696 | Park et al. | Jun 2013 | B2 |
8470388 | Zsinko et al. | Jun 2013 | B1 |
D685373 | Santos et al. | Jul 2013 | S |
8474994 | Campman | Jul 2013 | B2 |
8508928 | Killen et al. | Aug 2013 | B2 |
8547825 | Armstrong et al. | Oct 2013 | B2 |
8560501 | Deetz | Oct 2013 | B2 |
8566288 | Deetz | Oct 2013 | B1 |
8570738 | Fan et al. | Oct 2013 | B2 |
D694240 | Cho | Nov 2013 | S |
8583194 | Lio et al. | Nov 2013 | B2 |
8589643 | Meyer et al. | Nov 2013 | B2 |
D694758 | Muller | Dec 2013 | S |
8601201 | Scannell et al. | Dec 2013 | B2 |
8639849 | Achler | Jan 2014 | B2 |
D699212 | Tanaka | Feb 2014 | S |
D702650 | Reed et al. | Apr 2014 | S |
8701108 | Sekiguchi et al. | Apr 2014 | B2 |
8704771 | Shigeta | Apr 2014 | B2 |
D705836 | Huschka | May 2014 | S |
8715127 | Beutler et al. | May 2014 | B2 |
8717764 | Sun | May 2014 | B2 |
D706780 | Abdollahi et al. | Jun 2014 | S |
8782463 | Singh | Jul 2014 | B1 |
8793684 | Breitgand et al. | Jul 2014 | B2 |
D711382 | Eriksen | Aug 2014 | S |
8806480 | Araujo, Jr. et al. | Aug 2014 | B2 |
8832352 | Tsirkin et al. | Sep 2014 | B2 |
8839112 | Neerudu et al. | Sep 2014 | B2 |
8847797 | Barr | Sep 2014 | B1 |
8856255 | Ramakrishnan et al. | Oct 2014 | B2 |
8868860 | Beeken | Oct 2014 | B2 |
8880470 | Mashtizadeh et al. | Nov 2014 | B2 |
D719128 | Teulie | Dec 2014 | S |
D719571 | Sugilhara et al. | Dec 2014 | S |
D720353 | Nakamura | Dec 2014 | S |
D720357 | Kwon et al. | Dec 2014 | S |
8904383 | Bash et al. | Dec 2014 | B2 |
8909884 | Fleming et al. | Dec 2014 | B2 |
8930647 | Smith | Jan 2015 | B1 |
8938643 | Karmarkar et al. | Jan 2015 | B1 |
D722060 | Kwon et al. | Feb 2015 | S |
D722995 | Moreau | Feb 2015 | S |
8954645 | North et al. | Feb 2015 | B2 |
8964385 | Lam et al. | Feb 2015 | B2 |
D723554 | Choo et al. | Mar 2015 | S |
D723555 | Choo et al. | Mar 2015 | S |
8988821 | Li | Mar 2015 | B2 |
D726180 | Roat et al. | Apr 2015 | S |
9003144 | Hayes et al. | Apr 2015 | B1 |
D729242 | Kim et al. | May 2015 | S |
D729809 | Akana et al. | May 2015 | S |
D730307 | Tang et al. | May 2015 | S |
9035786 | Clifford et al. | May 2015 | B2 |
D731010 | Levitt et al. | Jun 2015 | S |
D731484 | Olsson et al. | Jun 2015 | S |
D733137 | Kawai | Jun 2015 | S |
9053068 | Tsirkin et al. | Jun 2015 | B2 |
D734754 | Ficacci | Jul 2015 | S |
D735726 | Chen | Aug 2015 | S |
9106020 | Cohen et al. | Aug 2015 | B2 |
9141625 | Thornewell et al. | Sep 2015 | B1 |
D740817 | Ignomirello | Oct 2015 | S |
D740818 | Ignomirello | Oct 2015 | S |
D741333 | Aoyagi et al. | Oct 2015 | S |
9170831 | Robinson et al. | Oct 2015 | B2 |
D742876 | Ignomirello | Nov 2015 | S |
D742883 | Akana et al. | Nov 2015 | S |
D742887 | Ignomirello | Nov 2015 | S |
D743404 | Ignomirello | Nov 2015 | S |
9183158 | O'Loughlin et al. | Nov 2015 | B2 |
9195401 | Li et al. | Nov 2015 | B2 |
9195491 | Zhang et al. | Nov 2015 | B2 |
9201674 | Rogel et al. | Dec 2015 | B2 |
9280380 | Tsirkin | Mar 2016 | B2 |
9304703 | Ignomirello | Apr 2016 | B1 |
9317326 | Ramanathan et al. | Apr 2016 | B2 |
9336042 | Brennerman et al. | May 2016 | B1 |
9336050 | Nakata et al. | May 2016 | B2 |
9354918 | Gupta et al. | May 2016 | B2 |
9377963 | Colbert et al. | Jun 2016 | B2 |
9389937 | Ackaret et al. | Jul 2016 | B2 |
9438534 | Xia et al. | Sep 2016 | B2 |
D768135 | Ignomirello | Oct 2016 | S |
D768136 | Ignomirello | Oct 2016 | S |
9467294 | Ignomirello | Oct 2016 | B2 |
9552217 | Tarasuk-Levin et al. | Jan 2017 | B2 |
9558005 | Sathyanarayana | Jan 2017 | B2 |
9582303 | Shu et al. | Feb 2017 | B2 |
9582319 | Ayala et al. | Feb 2017 | B2 |
9584312 | Ignomirello | Feb 2017 | B2 |
9603251 | Ignomirello | Mar 2017 | B1 |
9628108 | Ignomirello | Apr 2017 | B2 |
D786258 | Ignomirello | May 2017 | S |
9672054 | Gupta et al. | Jun 2017 | B1 |
9715347 | Ryu et al. | Jul 2017 | B2 |
9753768 | Chandrasekaran | Sep 2017 | B2 |
9766945 | Gaurav et al. | Sep 2017 | B2 |
9870834 | Li et al. | Jan 2018 | B2 |
10061514 | Ignomirello | Aug 2018 | B2 |
10120607 | Ignomirello | Nov 2018 | B2 |
10133636 | Ignomirello | Nov 2018 | B2 |
10224111 | Li et al. | Mar 2019 | B2 |
20020122597 | Kikuchi et al. | Sep 2002 | A1 |
20020172203 | Ji et al. | Nov 2002 | A1 |
20030018759 | Baumann | Jan 2003 | A1 |
20030115447 | Pham et al. | Jun 2003 | A1 |
20030122694 | Green | Jul 2003 | A1 |
20040017651 | Gan et al. | Jan 2004 | A1 |
20040143733 | Ophir et al. | Jul 2004 | A1 |
20040262409 | Crippen et al. | Dec 2004 | A1 |
20050091440 | Isa et al. | Apr 2005 | A1 |
20050099319 | Hutchison et al. | May 2005 | A1 |
20050191520 | Guntermann et al. | Sep 2005 | A1 |
20050198076 | Stata et al. | Sep 2005 | A1 |
20050210314 | Iguchi | Sep 2005 | A1 |
20050288800 | Smith et al. | Dec 2005 | A1 |
20060006517 | Lee et al. | Jan 2006 | A1 |
20060049956 | Taylor et al. | Mar 2006 | A1 |
20060171050 | Hanson et al. | Aug 2006 | A1 |
20060174140 | Harris et al. | Aug 2006 | A1 |
20060212644 | Acton et al. | Sep 2006 | A1 |
20060212692 | Ueno | Sep 2006 | A1 |
20060248273 | Jernigan, IV | Nov 2006 | A1 |
20060256866 | Ziauddin et al. | Nov 2006 | A1 |
20060285287 | Chen et al. | Dec 2006 | A1 |
20070081320 | Gilbert | Apr 2007 | A1 |
20070087586 | Karamooz | Apr 2007 | A1 |
20070101074 | Patterson | May 2007 | A1 |
20070164875 | Fredericks et al. | Jul 2007 | A1 |
20070180207 | Garfinkle | Aug 2007 | A1 |
20070206375 | Piepgras et al. | Sep 2007 | A1 |
20070236334 | Borovoy et al. | Oct 2007 | A1 |
20070291934 | Volkovs et al. | Dec 2007 | A1 |
20080005380 | Kawasaki et al. | Jan 2008 | A1 |
20080054845 | Wang | Mar 2008 | A1 |
20080059702 | Lu et al. | Mar 2008 | A1 |
20080062020 | Lakus-Becker | Mar 2008 | A1 |
20080077586 | Lam | Mar 2008 | A1 |
20080133835 | Zhu et al. | Jun 2008 | A1 |
20080291664 | Pesenti | Nov 2008 | A1 |
20080307128 | Amron | Dec 2008 | A1 |
20080310628 | Fujioka | Dec 2008 | A1 |
20090006640 | Brouwer | Jan 2009 | A1 |
20090040032 | Gerber | Feb 2009 | A1 |
20090055593 | Satoyama et al. | Feb 2009 | A1 |
20090061756 | Germagian | Mar 2009 | A1 |
20090100236 | Puig | Apr 2009 | A1 |
20090112880 | Oliveira et al. | Apr 2009 | A1 |
20090115646 | Duxbury | May 2009 | A1 |
20090119468 | Taylor et al. | May 2009 | A1 |
20090129691 | Luttmer et al. | May 2009 | A1 |
20090141435 | Kreiner et al. | Jun 2009 | A1 |
20090169001 | Tighe et al. | Jul 2009 | A1 |
20090193161 | Yoshida | Jul 2009 | A1 |
20090208221 | Sasai | Aug 2009 | A1 |
20090231153 | Hauenstein et al. | Sep 2009 | A1 |
20090300301 | Vaghani | Dec 2009 | A1 |
20090303709 | Weatherley | Dec 2009 | A1 |
20090307251 | Heller et al. | Dec 2009 | A1 |
20100076527 | Hammond et al. | Mar 2010 | A1 |
20100162039 | Goroff et al. | Jun 2010 | A1 |
20100169287 | Klose | Jul 2010 | A1 |
20100211737 | Flynn et al. | Aug 2010 | A1 |
20100245665 | Chrnega et al. | Sep 2010 | A1 |
20100250501 | Mandagere et al. | Sep 2010 | A1 |
20100253617 | Lio et al. | Oct 2010 | A1 |
20100257269 | Clark | Oct 2010 | A1 |
20100306294 | Schneider | Dec 2010 | A1 |
20110035361 | Soga | Feb 2011 | A1 |
20110107112 | Resch | May 2011 | A1 |
20110137519 | Christie | Jun 2011 | A1 |
20110170872 | Shin et al. | Jul 2011 | A1 |
20110276771 | Tajima et al. | Nov 2011 | A1 |
20110289294 | Maeda et al. | Nov 2011 | A1 |
20120076298 | Bolotov et al. | Mar 2012 | A1 |
20120124282 | Frank et al. | May 2012 | A1 |
20120131293 | Benhase et al. | May 2012 | A1 |
20120159282 | Ito | Jun 2012 | A1 |
20120166448 | Li et al. | Jun 2012 | A1 |
20120166757 | Volvovski et al. | Jun 2012 | A1 |
20120191982 | Levin | Jul 2012 | A1 |
20120209448 | Brower | Aug 2012 | A1 |
20120209559 | Brower | Aug 2012 | A1 |
20120239860 | Atkisson et al. | Sep 2012 | A1 |
20120246472 | Berengoltz | Sep 2012 | A1 |
20120278382 | Faith | Nov 2012 | A1 |
20120278553 | Mudhiganti et al. | Nov 2012 | A1 |
20120285738 | Cochrane et al. | Nov 2012 | A1 |
20120303359 | Mizuguchi et al. | Nov 2012 | A1 |
20120330894 | Slik | Dec 2012 | A1 |
20130013618 | Heller et al. | Jan 2013 | A1 |
20130019076 | Amidi et al. | Jan 2013 | A1 |
20130106297 | Yeh | May 2013 | A1 |
20130111166 | Resch et al. | May 2013 | A1 |
20130138764 | Satapathy | May 2013 | A1 |
20130211608 | Farrell et al. | Aug 2013 | A1 |
20130212161 | Ben-Shaul et al. | Aug 2013 | A1 |
20130268812 | Liu et al. | Oct 2013 | A1 |
20130283038 | Kulkarni et al. | Oct 2013 | A1 |
20140062307 | Spencer et al. | Mar 2014 | A1 |
20140082324 | Elhamias et al. | Mar 2014 | A1 |
20140119028 | Sato et al. | May 2014 | A1 |
20140146514 | Yahata | May 2014 | A1 |
20140211423 | Nguyen et al. | Jul 2014 | A1 |
20140223118 | Ignormirello | Aug 2014 | A1 |
20140223196 | Ignormirello | Aug 2014 | A1 |
20140258533 | Antony | Sep 2014 | A1 |
20140259014 | Watanabe et al. | Sep 2014 | A1 |
20140279911 | Ignormirello | Sep 2014 | A1 |
20140281257 | Hochberg et al. | Sep 2014 | A1 |
20140297938 | Puthiyedath et al. | Oct 2014 | A1 |
20140313043 | Heydron | Oct 2014 | A1 |
20140313700 | Connell et al. | Oct 2014 | A1 |
20140333449 | Thiesfeld et al. | Nov 2014 | A1 |
20150015405 | Bark et al. | Jan 2015 | A1 |
20150052282 | Dong | Feb 2015 | A1 |
20150058543 | Chan et al. | Feb 2015 | A1 |
20150084270 | Lo et al. | Mar 2015 | A1 |
20150095443 | Yang et al. | Apr 2015 | A1 |
20150098205 | Keranen et al. | Apr 2015 | A1 |
20150117019 | Kuenzler et al. | Apr 2015 | A1 |
20150131965 | Yoshioka | May 2015 | A1 |
20150163060 | Tomlinson et al. | Jun 2015 | A1 |
20150211916 | McGinn et al. | Jul 2015 | A1 |
20150212263 | Tzeng | Jul 2015 | A1 |
20150254092 | Chandrasekaran | Sep 2015 | A1 |
20150309839 | Lu | Oct 2015 | A1 |
20150324388 | Benke et al. | Nov 2015 | A1 |
20160011802 | Berke | Jan 2016 | A1 |
20160043484 | Brodsky et al. | Feb 2016 | A1 |
20160085022 | Yang et al. | Mar 2016 | A1 |
20160092203 | Filali-Adib et al. | Mar 2016 | A1 |
20160196158 | Nipane et al. | Jul 2016 | A1 |
20160217047 | Ignormirello et al. | Jul 2016 | A1 |
20160218737 | Ignomirello | Jul 2016 | A1 |
20160232839 | Ignomirello | Aug 2016 | A1 |
20160378547 | Brouwer et al. | Dec 2016 | A1 |
20170026172 | Ignomirello | Jan 2017 | A1 |
20170129572 | Mellor | May 2017 | A1 |
20170220498 | Ignomirello | Aug 2017 | A1 |
20170221322 | Ignomirello | Aug 2017 | A1 |
20180267865 | Ignomirello et al. | Sep 2018 | A1 |
Number | Date | Country |
---|---|---|
2327998 | Jun 2001 | CA |
2638672 | Feb 2009 | CA |
201426214 | Mar 2010 | CN |
102611560 | Jul 2012 | CN |
203327301 | Dec 2013 | CN |
0403232 | Dec 1990 | EP |
2393009 | Dec 2011 | EP |
2897018 | Jul 2015 | EP |
2407427 | Apr 2005 | GB |
H07-504541 | May 1995 | JP |
2000-31831 | Jan 2000 | JP |
2007-272826 | Oct 2007 | JP |
2007-281763 | Oct 2007 | JP |
2008-107615 | May 2008 | JP |
2008-165293 | Jul 2008 | JP |
2009-059096 | Mar 2009 | JP |
2009-524882 | Jul 2009 | JP |
2009-251725 | Oct 2009 | JP |
2010-512057 | Apr 2010 | JP |
2010072740 | Apr 2010 | JP |
2011-203842 | Oct 2011 | JP |
2012-027587 | Feb 2012 | JP |
2012-80470 | Apr 2012 | JP |
2012-129785 | Jul 2012 | JP |
WO 2006042019 | Apr 2006 | WO |
WO 2006042041 | Apr 2006 | WO |
WO 2010138824 | Dec 2010 | WO |
WO 2012117658 | Jul 2012 | WO |
WO 2014121102 | Aug 2014 | WO |
WO 2014170617 | Oct 2014 | WO |
WO 2015039181 | Mar 2015 | WO |
WO 2016168007 | Oct 2016 | WO |
WO 2017176523 | Oct 2017 | WO |
Entry |
---|
“CleanSync 2.0”, CleanSync Team, retireved from http://www.top4download.com/cleansync/xfliyozl.html, dated Feb. 25, 2012, 3 pages. |
“Disk Cloning and Mass System Deployment” retrieved from http://www.drive-image.com/Disk_Cloning_and_Mass_System_Deployment.shtml, 21 pages. |
“Enabling Memory Reliability, Availability, and Serviceability Features on Dell PowerEdge Servers,” http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.692.4199&rep=rep1&type=pdf, Dell Power Solutions, Aug. 2015. |
“IBM TotalStorage Enterprise Server Model 800”, IBM, retrieved from http://ps-2.kev009.com/rs6000/manuals/SAN/ESS/ESS800_Product_Data_sheet.pdf, dated 2003, 6 pages. |
“Lenovo Servers RAS Features,” https://youtu.be/zZyDxnCUE4c, Jan. 11, 2016. |
“Memory RAS technologies for HPE ProLiant/Synergy/Blade Gen10 servers with Intel Xeon scalable processors”, retrieved from https://www.hpe.com/h20195/V2/GetPDF.aspx/4AA4-3490ENW.pdf, Hewlett Packard Enterprises, Oct. 2017, 9 pages. |
“Ntfsclone-Efficiently clone, image, restore or rescue an NTFS”, Ubuntu, retrieved from http://manpages.ubuntu.com/manpages/gutsy/man8/ntfsclone.8.html, 4 pages. |
“Restore a System Image BackUp to Computer with Dissimilar Hardware”, AOMEI, retrieved from http://www.backup-utility.com/features/dissimilar-hardware-restore.html, 5 pages. |
“Revisiting Memory Errors in Large-Scale Production Data Centers: Analysis and Modeling of New Trends from the Field” by Justin Meza, Qiang Wu, Sanjeev Kumar, and Onur Mutlu, Carnegie Mellon University and Facebook, Inc., 2015 {Facebook fleet analysis over 14 months}. |
Alibaba, “new arrival fashion design zinc alloy metal U shap handle for leather of bag accessories qifeng f-994”, retrieved on Jul. 18, 2015 from http://alibaba.com/product-detail/new-arrival-fasion-design-zinc-alloy_1402892576.html?smToken=f4543babcb5440fda870ba4c2a621f1c&smSign=2q%Bp5jXRBjWkFnOu62%2Fy%2Fw%3D%3D, 1 page. |
Aliexpress, “F238F 3.5″ Hot-Swap SAS/SATA Hard Disk Drive Caddy/Tray”, retrieved on Jul. 14, 2015 from http://www.aliexpress.com/store/product/F238F-3-5-Hot-Swap-SAS-SATA-Hard-Disk-Drive-Caddy-Tray/1129178_1677183967.html, 1 Page. |
All Data Parts, “ICY DOCK ToughArmor MB992SK”, retrieved on Jan. 29, 2016 from http://www.alldataparts.com/index.php?main_page=product_info&cPath=59_699_714&products_id=65914&zenid=c7a2e4f22f81deb6482a7df9318ad922, 2 pages. |
American Standard Circuits, Inc., “High Temperature Organic Adhesive Bonding Films”, retrieved from http:// www.asc-i.com/preview/technology/thermal-managemenl/, on May 26, 2015, 5 pages. |
Anonymous: “Find Seed from Random Numbers”, Blitz Research Programming Forums, Dec. 4, 2011 (Dec. 4, 2011), XP055297586, retrieved from the Internet: http://web.archive.org/web/20111204053405/http://blitzbasic.com/Community/posts.php?topic=61025 (5 pages total). |
Anonymous: “Huffman Coding”, Wikipedia, the free encyclopedia, Jan. 18, 2013 (Jan. 18, 2013), pp. 1-10, XP055303125, Retrieved from the Internet: http://en.wikipedia.org/w/index.phptitle=Huffmancoding&oldid=533661878 [retrieved on Sep. 15, 2016] (10 pages total). |
Anonymous: “Zero suppression”, Wikipedia, the free encyclopedia, Jul. 25, 2012 (Jul. 25, 2012), XP055303438, Retrieved from the Internet: http://en.wikipedia.org/w/index.php?title=Zerosuppression&oldid=504088023 (retrieved on Sep. 16, 2016) (2pages total). |
Backblaze, “Backblaze Faceplates”, Sep. 8, 2015, retrieved on Jan. 29, 2016 from hllps://www.backblaze.com/blog/the-next-backblaze-face-plate/, 9 pages. |
CCLONLINE, “Cougar MX300 Gamer Case, ATX, No PSU, Max 7 Fans, 7 PCI Cards, USB 3.0, Honeycomb Front”, retrieved on Jul. 14, 2015 from http://www.cclonline.com/product/167472/MX300/Cases/COUGAR-MX300-Gamer-Case-ATX-No-PSU-Max-7-Fans-7-PCI-Cards-USB-3-0-Honeycomb-Front/CAS1679/, 2 pages. |
Chapter 15. “Cooling of Electronic Equipment”, retrieved from http://highered.mheducation.com/sites/dl/free/0073398128/835451/Chapter15.pdf, 70 pages. |
Checkline Europe, “ZT-DPU Digital Force Guage with luminescent display & remote sensor,” retrieved on Jul. 17, 2015 from http://www.checkline-eurpe.co.uk/digital_force_gaugesl/ZT-DPU, 6 pages. |
Chung et al., “Low Tg Epoxy Adhesives for Thermal Management”, retrieved from www.aitechnology.com/uploads/pdf/WHITE%20PAPER/lowTgEpoxy.pdf, 14 pages. |
Cisco, “10 Gigabit Ethernet Switching for High-Performance, Rack-Optimized Server Switiching”, retrived on Jul. 18, 2015 from http://www.cisco.com/c/en/us/products/collateral/switches/catalyst-4900-series-switches/product_data_sheet0900aecd80246552.html, 14 pages. |
Cisco, “Cisco Catalyst 4948 10 Gigabit Ethernet Switch Data Sheet”, retrieved on Jul. 20, 2015 from http://www.cisco.com/c/en/us/products/collateral/switches/catalyst-4900-series-switches/product_data_sheet0900aecd80246552.html, 14 pages. |
Cisco, “Cisco HDT Network Storage Hard Drive Trays Cisco Small Business Network Storage Systems”, retrieved on Jul. 20, 2015 from http://www.cisco.com/c/en/us/products/collateral/storage-networking/hdt-network-storage-hard-drive-trays/data_sheet_c78-502733.pdf, 3 pages. |
Cisco, “Replacing Cisco ASR 9000 Series Router Components”, retrieved on Jan. 29, 2016 from http://www.cisco.com/c/en/us/td/docs/routers/asr9000/hardware/installation/guide/asr9kHIGbk/asr9klGmaintaining.html, 15 pages. |
Cisco, “Replacing the Cisco 3745 Faceplate/Fan Tray”, manual, retrieved on Jul. 14, 2015 from http://www.cisco.com/public/scc/compass/3700/tasks-task_3745_faceplate_replace.htm, 1 Page. |
Colo Advisor, “High Density Colocation—Planning Pays Off”, retrieved on Jul. 18, 2015 from http://www.coloadvisor.com/blog/high-density-colocation-planning-pays-off/, 5 Pages. |
D. Tang et al., “Assessment of the Effect of Memory Page Retirement on System RAS Against Hardware Faults,” DSN, 2006. |
Delcom Products, “USB Visual Signal Indicator RGB”, retrieved on Apr. 4, 2016 from https://www.delcomproducts.com/productdetails.asp?prodcutnum=804000, 2 pages. |
Dell, “Hard Disk Drive Caddy or Tray”, retrieved on Jan. 29, 2016 from http://www.aliexpress.com/store/producl/F238F-3-5-Hot-Swap-SAS-SATA-Hard-Disk-Drive-Caddy-Tray/1129178_ 1677183967 .html, 4 Pages. |
Dell, “Installing Enclosure Components”, retrieved on Jul. 18, 2015 from http://ftp.respmech.com/pub/MD3000/en/ MD3000_0wners_Manual/install.htm, 11 Pages. |
Dhgate, 120mm Apexi 3 in 1 Guage hu's Store Dhgate, retrieved on Jul. 17, 2015 from http://webcache.googleusercontent.com/search?q=cache:YpQK9gy7APQJ:www.dhgate.com/store/product/120mm-apexi-3-in-1-guage-tachometer-rpm-water/141993290.html+&cd=6&hl=en&ct=clnk&gl=in, 3 pages. |
Dictionary Techniques, C.M. Liu, National Chiao-Tung University Website, Dec. 31, 2010, website: http://people.cs.nctu.edu.tw/˜cmliu/Courses/Compression/chap5.pdf, 137 pages. |
DRAM errors in the wild: a large-scale field study, Bianca Schroeder, Eduardo Pinheiro, Wolf-Dietrich Weber, SIGMETRICS/Performance'09, Jun. 15-19, 2009, Seattle, WA, USA {largescale Google study for 2.5 years}. |
Ebay, “Dell PowerEdge M610 Server 2XQUAD Core 2.8GHz/X5560 64GB 2X300GB Raid Rails”, retrieved on Jul. 18, 2015 from http://www.ebay.ca/itm/271899179228?rmvSB=true, 3 pages. |
Ebay, “Vintage VGC Chrome Chevron Boomerang Drawer Cabinet Door Pulls Washington”, retrieved on Jul. 18, 2015 from http://www.ebay.com/itmNintage-VGC-CHROME-Chevron-Boomerang-DRAWER-Cabinet-Door-Pulls-Washington/200818869564? trksid=p204 7675.c100005.m1851&_trkparms=aid%3D222007%26algo%3DSIC.MBE% D26ao%3D1%26asc%3D32298%26meid%3D3280c5c9af794a00afd5a 1 a69fc33f59%26pid%3D, 1 Page. |
Edwards Signaling, “125XBRi Series XTRA-BRITE”, Data Sheet ES001-0114, No. 1, retrieved on Apr. 4, 2016 from https://edwards-signals.com/files/125XBRI_Series_datasheet_v10.pdf, 2 pages. |
European Search Report dated Sep. 5, 2016 in connection with corresponding EP Patent Application No. 16165443.9 (9 pages total). |
Extended European Search Report dated Sep. 19, 2016 issued in connection with corresponding EP Application No. 14745861.6 (9 pages total). |
Extended European Search Report dated Aug. 10, 2016 issued in connection with corresponding EP Application No. 14745756.8 (9 pages total). |
EZ-PC, “Infrastructure and Networking”, retrieved on Jul. 18, 2015 from http://www.ez-pc.org/services/infrastruture, 3 pages. |
Frostytech, “Copper Honeycomb Concept Heatsink—BBHS Thermal Solutions Corp”, retrieved on Jul. 14, 2015 from http://www.frostytech.com/permalinkArch.cfm?NewsID=88219, 1 Page. |
Gilbert, “The Bus (PCI and PCI-Express)”, retrieved from hppt://www.yale.edu/pclt/PCHW/BUS.HTM on Apr. 28, 2015 dated Jan. 4, 2008, 5 pages. |
Hague Registration No. DM/056877, retrieved on Jul. 18, 2015 from http://www.wipo.int/designdb/hague/en/showData.jsp?SOURCE=HAGUE&KEY=DM056877, 3 page. |
Hardware Secrets, “Everything You Need to Know About the PCI Express”, retrieved from http://www.hardwaresecrets.com/article/EverythingYouNeedtoKnowAboutthePCIExpress/190 on Jul. 12, 2012, 3 pages. |
Hewlett Packard (HP), “Best Practices for HP BladeSystem Deployments using HP Serviceguard Solutions for HP-UX 11i”, HP Publication No. 5697-0470, May 2010, pp. 1-26. |
Highly Relable Systems, “RAIDFrame DAS 5000”, retrieved on Jul. 18, 2015 from https://www.high-rely.com/products/raidframe5bay/1 page. |
Highly Reliable Systems, “MPac-G2 Media Tray”, retrieved on Jan. 29, 2016 from https:/Iwww.high-rely.com/products/mpac-media/, 1 page. |
Highly Reliable Systems, “RAID Frame 5 Bay”, retrieved on Jan. 29, 2016 from https:/lwww.high-rely.com/products/raidframe5bay/, 1 page. |
IDOTPC, “BR-0804 4pc/set Rackmount Half-1 U Joint Plate Kit”, retrieved on Jul. 18, 2015 from http://www.idotpc.com, thestore/pc/BR-0804-4pc-set-Rackmount-Half-1U-Joint-Plate-Kit-p1343.htm, 1 Page. |
Intel, “Intel 64 and IA-32 Architectures Optimization Reference Manual”, Manual, Sep. 2014, 642. |
International Computer Concepts Inc, “PCI & PCIe Overview/Comparision”, retrieved from http://www.iccusa.com/comparepci/ on Apr. 28, 2015, 5 pages. |
International Preliminary Report on Patentability dated Aug. 13, 2015 issued in connection with International Application No. PCT/US2014/014209, 8 pages. |
International Preliminary Report on Patentability dated Aug. 13, 2015 issued in connection with International Application No. PCT/US2014/014225, 7 pages. |
International Search Report and Written Opinion dated Apr. 14, 2017, issued in connection with corresponding International Application No. PCT/US2017/015544 (13 pages total). |
International Search Report and Written Opinion dated Aug. 20, 2014 issued in connection with International Application No. PCT/US2014/014209, 5 pages. |
International Search Report and Written Opinion dated Jan. 16, 2015 issued in connection with International Application No. PCT/US2014/014225, 2 pages. |
International Search Report for International Application No. PCT/US17/24692 dated Jul. 27, 2017. |
Japanese Office Action dated Sep. 20, 2016, issued in connection with corresponding Japanese Patent Application No. 2016-080683 with English Language translation (8 pages total). |
Japanese Office Action dated Mar. 14, 2017, issued in connection corresponding JP Application No. 2016-080683 (9 pages total). |
Koike et al. Analyse validity of dedup technology, Information Processing Association Reports, 2011, 6, “DVD-ROM”, Japan, General Legal Entity, Information Processing Society, May 1, 2012, vol. 2012-EVA-37, No. 1, pp. 1 to 6. |
Lillibridge, M., Eshghi, K., Bhagwat, D., Deolalikar, V., Trezise, G., and Camble, P. 2009. Sparse indexing: Large scale, inline deduplication using sampling and locality. In Proceedings of the 7th USENIX Conference on File and Storage Technologies (FAST). USENIX Association, Berkely, CA, 111-123. |
Mactech, “NES Cartridge External Hard Drive”, retrieved on Jul. 18, 2015 from http://www.mactech.com/category/ type-article/macmod-web-site/miscellaneous-mods, 2 Pages. |
Makatos, Thano, et al., ZBD: Using Transparent Compression at the Block Level to Increase Storage Space Efficiency, IEEE Computer Society pp. 61-70 (2010). |
McCabe, “How to Add a Hard Drive to Your Windows Home Server”, Information Week Network Computing, retrieved on Jul. 14, 2015 from http://www.networkcomputing.com/storage/how-to-add-a-hard-drive-to-your-windows-home-server-/d/d-id/1098397?, 3 Pages. |
Oracle Corp, “I/O Domains and PCI Express Buses”, retrieved from docs.oracle.com/cd/E1960401/8210406/configurepciexpressbusesacrossmultipledoms/index.html on Apr. 28, 2015, 3 pages. |
Philips, “Philips AZ302/98 CD Sound Machine”, retrieved on Feb. 5, 2016 from http://www.naaptol.com/portable-speakers/philips-az302/p/641367.html, 3 pages. |
Saduqulla et al., “Threshold Proxy Re-Encryption in Cloud Storage System, International Journal of Advanced Research in Computer Science and Software Engineering”, vol. 3, Issue 11, retrieved from http://www.ijarcsse.com/docs/papers/Volume_3/11_November2013/V3/November0332.pdf, dated Nov. 2013, 5 pages. |
Samsung Memory DDR4 SDRAM, http://www.samsung.com/us/dell/pdfs/DDR4_Brochure_2015.pdf, Samsung Electronic Co., 2 pages, 2015. |
Seagate, SCSI Commands Reference Manual (Feb. 14, 2006). |
Smith, Steven W. Data Compression Tutorial: Part 1,: EE Times, Jun. 14, 2007. |
Startech, “Hot Swap Hard Drive Tray for SATSASBAY3BK”, retrieved on Feb. 1, 2016 from http://www.startech.com/HDD/Mobile-Racks/Extra-25inch-or-35inch-Hot-Swap-Hard-Drive-Tray-for-SATSASBAY3BK˜SATSASTRAYBK, 3 pages. |
Sun Microsystems, Best Practices for Sun StorEdge 6920 System (Nov. 2004). |
Symbolic IO Corporation, International Search Report and Written Opinion dated May 3, 2016, for PCT/US16/25988, 8 pages. |
Tate, Jon, et al., Real-time Compression in SAN Volume Controller and the Storwize V 7000, IBM (Aug. 17, 2012). |
Weil, Sage A., CRUSH: Controlled, Scalable, Decentralized Placement of Replicated Data; IEEE—2006; p. 1-12. |
Wylie et al., “Selecting the right data distribution scheme for a survivable storage system”, CMU-CS-01-120, Carnegie Mellon University, retrieved from http://www.pdl.cmu.edu/PDL-FTP/Storage/CMU-CS-01-120.pdf, dated May 2001, 23 pages. |
Xilinx, Inc., Virtex-5 Family Overview, Feb. 6, 2009, 13 pages. |
“Elastic Online Analytical Processing on RAMCloud4,” Tinnefeld et al., EDBT/ICDT (Mar. 18-22, 2013), available at https://openproceedings.org/2013/conf/edbt/TinnefeldKGBRSP13.pdf. |
“RamCloud Project”, accessed Apr. 3, 2018, available at https://ramcloud.atlassian.net/wiki/spaces/RAM/overview. |
“Breakthrough Nonvalatile Memory Technology,” Micron, 3D Xpoint Technology, accessed Apr. 3, 2018, available at https://www.micron.com/products/advanced-solutions/3d-xpoint-technology. |
Number | Date | Country | |
---|---|---|---|
20190114087 A1 | Apr 2019 | US |
Number | Date | Country | |
---|---|---|---|
62148160 | Apr 2015 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 15783838 | Oct 2017 | US |
Child | 16039650 | US | |
Parent | PCT/US2016/025988 | Apr 2016 | US |
Child | 15783838 | US | |
Parent | 14804175 | Jul 2015 | US |
Child | PCT/US2016/025988 | US |