Intel Corporation

Reshma Lal

Pradeep M Pappachan

Siddhartha Chhabra

Luis S Kida

Soham Jayesh Desai

Bin Xing

Steven Bernard McGowan

Alpa Narendra Trivedi

Vincent R Scarlata

Sujoy Sen

Selvakumar Panneer

Robert O Sharp

Krystof C Zmudzinski

Abhishek Basak

Victoria C Moore

Anna Trikalinou

Salessawi Ferede Yitbarek

Ned M Smith

Lawrence A Booth

Gideon Gerzon

Brent D Thomas

Matthew Eldon Hoekstra

Reouven Elbaz

Srikanth Varadarajan

Vedvyas Shanbhogue

David J Harriman

Kapil Sood

Santosh Ghosh

Ioannis T Schoinas

Raghunandan Makaram

Baruch Chaikin

Kumar Narasimhan Dwarakanath

Yu-Yuan Chen

Rakesh Avichal Ughreja

Micah James Sheller

William A Stevens, Jr

Akshay Ramesh Kadam

Sarbartha Banerjee

Prashant Dewan

Baiju V Patel

Utkarsh Y Kakaiya

Jason Martin

Daniel Nemiroff

Nathan Heldt-Sheller

Ronald Perez

Jianping Xu

David M Lee

Jeffrey Sedayao

Ivan Jibaja

Bharat S Pillilli

Anahit Tarkhanyan

Kar Leong Wong

Eric Innis

Christine E Severns-Williams

Adeel A Aslam

Anjo Lucas Vahldiek-Oberwagner

Michael Steiner

Saravana Priya Ramanathan

Sudha Krishnakumar

Nanda Kumar Unnikrishnan

Marcin Andrzej Chrapek

Marcela S Melara

Ravi L Sahita

Rajesh Poornachandran

Alex Nayshtut

Dheeraj R Subbareddy

Omer Ben-Shalom

Edwin Verplanke

Anjali Singhai Jain

Ulhas S Warrier

Bo A Zhang

Hakan Magnus Eriksson

Joseph J Grecco

Mark Shanahan

Josh Triplett

Mrittika Ganguli

Ismo Puustinen

Chih-Jen Chang

Abhirupa Layek

Priya Autee

Mikko Ylinen

David Hines

Shubha Bommalingaiahnapallya

Gaurav Kumar

Pallavi Dhumal

Claire Vishik

Nupur Jain

Rick Edgecombe

Travis W Peters

Luis S Kida Kida

Richard Edgecombe

Mukesh Gangadhar Bhavani Venkatesan

Self-healing networks of Infrastructure Processing Units (IPUs) and associated methods and apparatus. The self-healing IPUs manage other processing units (XPU) clusters by seamlessly migrating the IPU responsibilities to another IPU in the networked environment (e.g., data center) that may be available when an IPU failures or becomes unavailable. A central Resource Manager is used to monitors the health of the IPUs in the data center and in the event of in IPU failure, locates another IPU and assigns it to take over the failed IPU's functions. Replacement and workload migration of a failed XPU in an IPU managed XPU cluster with a remote XPU that is network connected is also supported. The IPU monitors the health of the XPUs in its cluster an informs the Resource Manager of an XPU failure which locates another XPU in the data center and assigns it to the cluster that has the failed XPU.

Self-healing network of infrastructure processing units

Technologies for secure I/O include a compute device having a processor, a memory, an input/output (I/O) device, and a filter logic. The filter logic is configured to receive a first key identifier from the processor, wherein the first key identifier is indicative of a shared memory range includes a shared key identifier range to be used for untrusted I/O devices and receive a transaction from the I/O device, wherein the transaction includes a second key identifier and a trust device ID indicator associated with the I/O device. The filter logic is further configured to determine whether the transaction is asserted with the trust device ID indicator indicative of whether the I/O device is assigned to a trust domain and determine, in response to a determination that the transaction is not asserted with the trust device ID indicator, whether the second key identifier matches the first key identifier.

Technologies for filtering memory access transactions received from one or more I/O devices

An apparatus to facilitate disaggregated computing for a distributed confidential computing environment is disclosed. The apparatus includes one or more processors to: provide a remote GPU middleware layer to act as a proxy for an application stack on a client platform that is separate from the remote server platform, wherein the remote GPU middleware layer comprises is to expose an abstraction of the remote GPU to userspace components of a remote GPU stack, the userspace components running on the client machine; communicate with a kernel mode driver of the one or more processors to cause the host memory to be allocated for data structures used to communicate commands between the client and the remote GPU; and invoke the kernel mode driver to submit a workload generated by the application stack, the workload submitted for processing by the remote GPU using the data structures allocated in the host memory.

Disaggregated computing for distributed confidential computing environment

Technologies for secure I/O data transfer with an accelerator device include a computing device having a processor and an accelerator. The processor establishes a trusted execution environment. The trusted execution environment may generate an authentication tag based on a memory-mapped I/O transaction, write the authentication tag to a register of the accelerator, and dispatch the transaction to the accelerator. The accelerator performs a cryptographic operation associated with the transaction, generates an authentication tag based on the transaction, and compares the generated authentication tag to the authentication tag received from the trusted execution environment. The accelerator device may initialize an authentication tag in response to a command from the trusted execution environment, transfer data between host memory and accelerator memory, perform a cryptographic operation in response to transferring the data, and update the authentication tag in response to transferrin the data. Other embodiments are described and claimed.

Technologies for secure I/O with accelerator devices

An apparatus includes a central processing unit (CPU), including a plurality of processing cores, each having a cache memory, a fabric interconnect coupled to the plurality of processing cores and cryptographic circuitry, coupled to the fabric interconnect including mesh stop station to receive memory data and determine a destination of the memory data and encryption circuitry to encrypt/decrypt the memory data based on a destination of the memory data.

Cache optimization mechanism

A computing platform is disclosed. The computing platform includes a first computer system comprising a first graphics processing unit (GPU), a network coupled to the first computer system and a second computer system, coupled to the first computer system via the network, comprising a second GPU, wherein the first and second computer system are configured to perform distributed processing of graphics workloads between the first GPU and the second GPU.

Dynamic resource management mechanism

Embodiments are directed to trusted local memory management in a virtualized GPU. An embodiment of an apparatus includes one or more processors including a trusted execution environment (TEE); a GPU including a trusted agent; and a memory, the memory including GPU local memory, the trusted agent to ensure proper allocation/deallocation of the local memory and verify translations between graphics physical addresses (PAs) and PAs for the apparatus, wherein the local memory is partitioned into protection regions including a protected region and an unprotected region, and wherein the protected region to store a memory permission table maintained by the trusted agent, the memory permission table to include any virtual function assigned to a trusted domain, a per process graphics translation table to translate between graphics virtual address (VA) to graphics guest PA (GPA), and a local memory translation table to translate between graphics GPAs and PAs for the local memory.

Trusted local memory management in a virtualized GPU

Methods, apparatuses and system provide for technology that interleaves a plurality of verification commands with a plurality of copy commands in a command buffer, wherein each copy command includes a message authentication code (MAC) derived from a master session key, wherein one or more of the plurality of verification commands corresponds to a copy command in the plurality of copy commands, and wherein a verification command at an end of the command buffer corresponds to contents of the command buffer. The technology may also add a MAC generation command to the command buffer, wherein the MAC generation command references an address of a compute result.

Graphics security with synergistic encryption, content-based and resource management technology

Attestation of operations by tool chains is described. An example of a storage medium includes instructions for receiving source code for processing of a secure workload of a tenant; selecting at least a first compute node to provide computation for the workload; processing the source code by an attestable tool chain to generate machine code for the first compute node, including performing one or more conversions of the source code by one or more convertors to generate converted code and generating an attestation associated with each code conversion, and receiving machine code for the first compute node and generating an attestation associated with the first compute node; and providing each of the attestations from the first stage and the second stage for verification.

Attestation of operations by tool chains

An apparatus to facilitate metrics and security-based accelerator service rescheduling and auto-scaling using a programmable network device is disclosed. The apparatus includes processors to collect metrics corresponding to communication links between microservices of a service managed by a service mesh; determine, based on analysis of the metrics, that a workload of the service can be accelerated by offload to a hardware accelerator device; generate a scaling request to cause the hardware accelerator device to be allocated to a cluster of hardware devices configured for the service; cause the scaling request to be transmitted to a programmable network device managing the hardware accelerator device, the programmable network device to allocate the hardware accelerator device to the cluster and to register the hardware accelerator device with the service mesh; and schedule the workload of the service to the hardware accelerator device.

Metrics and security-based accelerator service rescheduling and auto-scaling using a programmable network device

An apparatus comprising translator circuitry to receive a plurality of physical addresses of memory data, determine an offset associated with each of the physical page addresses and apply a tweak seed to each offset to generate a plurality of tweaks.

Data transfer encryption mechanism

An apparatus to facilitate in-place memory copy during remote data transfer in a heterogeneous compute environment is disclosed. The apparatus includes a processor to receive data via a network interface card (NIC) of a hardware accelerator device; identify a destination address of memory of the hardware accelerator device to write the data; determine that access control bits of the destination address in page tables maintained by a memory management unit (MMU) indicate that memory pages of the destination address are both registered and free; write the data to the memory pages of the destination address; and update the access control bits for memory pages of the destination address to indicate that the memory pages are restricted, wherein setting the access control bits to restricted prevents the NIC and a compute kernel of the hardware accelerator device from accessing the memory pages.

In-place memory copy during remote data transfer in heterogeneous compute environment

Examples described herein relate to network layer 7 (L7) offload to an infrastructure processing unit (IPU) for a service mesh. An apparatus described herein includes an IPU comprising an IPU memory to store a routing table for a service mesh, the routing table to map shared memory address spaces of the IPU and a host device executing one or more microservices, wherein the service mesh provides an infrastructure layer for the one or more microservices executing on the host device; and one or more IPU cores communicably coupled to the IPU memory, the one or more IPU cores to: host a network L7 proxy endpoint for the service mesh, and communicate messages between the network L7 proxy endpoint and an L7 interface device of the one or more microservices by copying data between the shared memory address spaces of the IPU and the host device based on the routing table.

Network layer 7 offload to infrastructure processing unit for service mesh

An apparatus to facilitate protecting data transfer between a secure application and networked devices is disclosed. The apparatus includes a processor to provide a trusted execution environment (TEE) to run an application, wherein the processor is to: generate, via the application in the TEE, encrypted data, wherein the encrypted data comprises a payload; copy, via the application in the TEE, the encrypted data to a local buffer; interface, using the application in the TEE, with a source network interface controller (NIC) to initiate a copy over a network of the encrypted data from the local buffer to a remote buffer of a remote platform; and communicate, after completing the copy of the network of the encrypted data, at least one message with the remote platform to indicate that the encrypted data is available and to enable the remote platform to verify integrity of the encrypted data.

Protecting data transfer between a secure application and networked devices

Embodiments are directed to providing integrity-protected command buffer execution. An embodiment of an apparatus includes a computer-readable memory comprising one or more command buffers and a processing device communicatively coupled to the computer-readable memory to read, from a command buffer of the computer-readable memory, a first command received from a host device, the first command executable by one or more processing elements on the processing device, the first command comprising an instruction and associated parameter data, compute a first authentication tag using a cryptographic key associated with the host device, the instruction and at least a portion of the parameter data, and authenticate the first command by comparing the first authentication tag with a second authentication tag computed by the host device and associated with the command.

Integrity protected command buffer execution

Growing community of inventors

Hillsboro, OR, United States of America

Reshma Lal