Since 1986 - Covering the Fastest Computers
in the World and the People Who Run Them
February 01, 2012
Tilera Chips In with 64-bit Manycore Processors
SeaMicro's is not the only vendor stirring up the microserver space this week. Chipmaker Tilera this week announced its low-power 64-bit processors that are essentially aimed at the same scale-out datacenter market SeaMicro is targeting with its new Xeon-based microserver and older Atom-based gear. Those applications include ones running PHP and Java, in memory and NoSQL database workloads, and data mining apps based on Hadoop.
The two processors Tilera announced this week are the TILE-Gx16 and the TILE-Gx3, which are 16-core and 36-core versions, respectively. (The company says 64-core and 100-core implementations are in the pipeline.) The new chips are available in various frequencies between 1.0 and 1.5 GHz and are equipped with two integrated DDR3 memory controllers, 8 lanes of PCIe 2.0 connectivity, and multiple GigE/10GigE ports. Specific power draw on the various TILE-Gx parts has not been specified, although the company says to expect between 20 and 48 watts for "typical applications."
Like SeaMicro, Tilera uses on-chip integration to reduce the number of components needed on the motherboard, but the similarities stop there. Whereas SeaMicro uses Intel Atom or Xeon CPUs and its custom interconnect fabric to build a cluster in a box, Tilera has essentially built a cache-coherent cluster on a chip with an on-chip mesh network.
Also unlike SeaMicro, Tilera must count on server and appliance makers to incorporate their processors into standalone systems. Theoretically SeaMicro itself could build a server around Tilera silicon, and would almost certainly do so if there was enough demand.
As of today, Tilera says it has 20 customers in the pipeline for the new TILE-Gx products, with Harmonic and Mercury Computer Systems announced this week. The company promises more TILE-Gx-based boards and systems will be announced later in the year.
Discussion
There are 0 discussion items posted.
|
Join the Discussion |
Sponsored Links
DevDeck Developer Workstations offer a complete tool suite helping scientific programmers through the application development lifecycle for programming on existing and future many-core targets including CPU, GPU, and Intel MIC.
- Appro Xtreme-X™, SDSC’s Gordon Supercomputer Ready for Researchers
- Micro-Virtualization: Low-Fat Virtualization for High Performance
- Push back your limits with new generation bullx B510 blades
- Convey Launches next-generation system: 2-3X Performance Increase
- Altair Knows HPC
- SGI NAS – a complete modular open storage solution
- Webinar: Mellanox FDR 56Gb/s InfiniBand Switch Solutions
- 2X in 4 Weeks. Guaranteed.
- Accelerate Your Applications with Less Programming Effort
- Finding Hard-to-Reproduce Bugs with Reverse Debugging
- Get More ROI from your HPC. Sign up to “Test Drive” the ICE 8400.
- How do you make Grid Engine run faster?
RSS Feeds
Feeds by Topic
- Applications
- Developer Tools
- Interconnects
- Middleware
- Networks
- Processors
- Storage
- Systems
- Visualization
Feeds by Industry
Feeds by Content Type
- Features
- Blogs
- Around the Web
- Off the Wire
- Whitepapers
- Multimedia
- HPC Soundbite Podcasts
- HPCwire Job Bank
Subscribe to All Content
May 23, 2012
- Chelsio Releases Storage Software, Enables Scaling for iSCSI Devices
- LSU Paper Receives High Honors at International ACM Symposium
May 22, 2012
- Cornell University Updates BioInformatics Software
- Allinea Software Releases CUDA Education Pack
- PRACE Awarded 120 Million Core Hours on Hermit Supercomputer
- PRACE to Hold Summer School for Multi-Core, MIC Code Optimization
- Spectra Logic to Provide Tape Storage for Blue Waters Supercomputer
May 21, 2012
- UK Researchers Develop 'Super-Fast' Memory Chip
- Winners of Watson Academic Case Competition Announced
- Numerix Analytics Portfolio Now Available on Windows Azure
- HPC User Forum to be Held in London, Stuttgart
- Ciena Transferring 100G Traffic to Iceland with Single Fiber Pair
- Mellanox Creates Ethernet to InfiniBand Interconnect for HP Servers
May 18, 2012
- University of Southampton Builds UK's Largest Academic Supercomputer
- DataDirect Networks to Support Storage on Open Compute Project Hardware
May 17, 2012
- Part-Time Scientists to Deliver Lunar Rover Keynote at GTC
- NextIO Closes Series F Funding Round
- Dell Donated Cluster to Support Clinical Trial for Pediatric Cancer
- Rice University Develops Low Power, Inexact Chip
May 16, 2012
- DOD Makes Multi-System Supercomputing Acquisition Totaling Four Petaflops
- Mercury Computer Systems Develops Sensor Stream Delivery Application
- HPC Project Updates Par4All Parallelization Platform
- Limit Point Systems, MBA Sciences Collaborate to Assist Oil and Gas Industry
- Nvidia Announces Availability of Tesla K10 GPU
- GE Announces Graphics Board Based on Nvidia VPX Platform
- DOE INCITE Program to Allocate Titan, Mira Compute Resources
May 15, 2012
- SGI Adds Kepler GPU-Accelerated Servers to Rackable Line
- AMAX Intros HPC Servers with New Kepler GPUs
- Nvidia Introduces GPU Virtualization Platform
- Wolfram Releases Financial Platform
- Dell Adds Quad-Socket Xeon Server to Lineup
- Nvidia Announces Kepler-Based Tesla GPUs
- OpenFabrics Alliance, Microsoft Bring RDMA to Windows Server 2012
- Penguin Computing Adds New Sandy Bridge Chips to Servers
- Caltech Research Center Using Panasas Parallel Storage
May 14, 2012
- BOXX Workstations Featuring NVIDIA Maximus Demo at GPU Technology Conference
- Registration Now Open for XSEDE12
- Blue Waters Supercomputer Debugs With Allinea DDT
- Appro Xtreme-X Supercomputer Powered by New Intel Processors
- AMAX Releases Systems based on Intel E3 Family
- IBM Augments x86 Offerings
- Super Micro Shipping Servers With New Sandy Bridge Processors
- HP Updates Gen8 Servers With New Xeon Families
- SGI Confirms Support for New Xeon Chips
- Intel Adds More Chips to Xeon Lineup
- BOXX Technologies Demos Workstations at GTC
- nCore Design to Host Multicore Programming Course
- Mellanox Releases InfiniBand Driver for VMWare
- NVIDIA Intros GPU Application Development Environment for Linux, Mac-OS
May 11, 2012
May 10, 2012
- EMC Acquires XtremIO
- National Computational Science Institute, XSEDE Offer Workshops to Educators
- Rogue Wave Acquires IBM Visualization Products
- Kove Storage Claims Sustained Low Latency Performance
Most Read Features
- Thomas Sterling: 'I Think We Will Never Reach Zettaflops'
- TACC-Intel Symposium Highlights MIC Architecture Developments
- Intel Makes a Deal for Cray's Interconnect Assets
- NVIDIA Launches Kepler Into HPC
- Novel Chip Technology to Power GRAPE-8 Supercomputer
- Intel Rolls Out New Server CPUs
- Some Thoughts on Intel’s Acquisition of Cray's Interconnect Technology
- ESnet Launches Architecture to Help Researchers Deliver on Data-Intensive Science
- OpenACC Starts to Gather Developer Mindshare
- The Last Mile of Virtualization
Most Read Around the Web
- Famous Physicist Predicts the End of Silicon-Based Computing
- Zettaflops Will Happen Says HPC Analyst
- NVIDIA Launches Petaflop App Contest
- Chips on the Table: Sandy Bridge versus Westmere
- Tiny Crystal Turns Out to be a Quantum Computing Gem
- Designing Faster Chips with... Faster Chips
- Next-Gen Memory on the Horizon
- China Looks to a National Processor Architecture
- NVIDIA Unveils Virtualized GPU with Kepler-Based Board
- US Energy Secretary Talks Supercomputing
Most Read This Just In
- Open Source Compiler Adds Support for Nvidia GPUs
- Cray to Sell Interconnect Hardware, IP to Intel
- NextIO Closes Series F Funding Round
- Foundation to Fund Institute for Theoretical Computer Science
- Open Fabrics Recommends Linux-Specific Distributions of RDMA Middleware
- Chinese Genomics Giant Launches GPU-Accelerated Cloud Service for DNA Sequencing
- DOD Makes Multi-System Supercomputing Acquisition Totaling Four Petaflops
- GE Announces Graphics Board Based on Nvidia VPX Platform
- NetApp Updates Storage Platform
- University of Southampton Builds UK's Largest Academic Supercomputer
Most Read Blogs
Feature Articles
NVIDIA Works On CPU Co-Dependency Issues with Kepler GPU
NVIDIA is telling everyone that the GK110, its new Kepler GPU aimed at supercomputing, is all about improving performance per watt. But the other driving theme behind the new architecture is reducing the GPU's reliance on its CPU host. How well it accomplishes both these goals areas could determine the success of the new chip in high performance computing.
Read more...
OpenACC Starts to Gather Developer Mindshare
PGI, Cray, and CAPS enterprise are moving quickly to get their new OpenACC-supported compilers into the hands of GPGPU developers. At NVIDIA's GPU Technology Conference this week, there was plenty of discussion around the new HPC accelerator framework, and all three OpenACC compiler makers, as well as NVIDIA, were talking up the technology.
Read more...
NVIDIA Launches Kepler Into HPC
NVIDIA has introduced its first Kepler-generation GPU product for high performance computing, and revealed some of the inner working of the new architecture. The announcement took place at the kickoff of the company's GPU Technology Conference taking place this week in San Jose, California.
Read more...
Sponsored Whitepapers
Sponsored Multimedia
Newsletters
HPC Job Bank
Featured Events
-
May 21, 2012
- May 25, 2012
IPDPS 2012 - 26th IEEE International Parallel & Distributed Processing Symposium
Shanghai,
China
-
June 17, 2012
- June 21, 2012
ISC '12 - The HPC Event
Hamburg,
Germany
-
June 17, 2012
- June 17, 2012
HPC Advisory Council European Conference 2012
Hamburg,
Germany
-
June 25, 2012
- June 29, 2012
International Advanced Computing Workshop on High Performance Computing, Grid and Clouds
Centraro,
Italy
-
September 17, 2012
- September 21, 2012
EGI Technical Forum 2012
Prague,
Czech Republic
