Intel Skylake-X and Skylake-SP Utilize Mesh Architecture for Intra-Chip Communication

Subject: Processors | June 15, 2017 - 04:00 PM |
Tagged: xeon scalable, xeon, skylake-x, skylake-sp, skylake-ep, ring, mesh, Intel

Though we are just days away from the release of Intel’s Core i9 family based on Skylake-X, and a bit further away from the Xeon Scalable Processor launch using the same fundamental architecture, Intel is sharing a bit of information on how the insides of this processor tick. Literally. One of the most significant changes to the new processor design comes in the form of a new mesh interconnect architecture that handles the communications between the on-chip logical areas.

Since the days of Nehalem-EX, Intel has utilized a ring-bus architecture for processor design. The ring bus operated in a bi-directional, sequential method that cycled through various stops. At each stop, the control logic would determine if data was to be the collected to deposited with that module. These ring bus stops are located at memory controllers, CPU cores / caches, the PCI Express interface, memory controllers, LLCs, etc. This ring bus was fairly simple and easily expandable by simply adding more stops on the ring bus itself.

xeon-processor-5.jpg

However, over several generations, the ring bus has become quite large and unwieldly. Compare the ring bus from Nehalem above, to the one for last year’s Xeon E5 v5 platform.

intel-xeon-e5-v4-block-diagram-hcc.jpg

The spike in core counts and other modules caused a ballooning of the ring that eventually turned into multiple rings, complicating the design. As you increase the stops on the ring bus you also increase the physical latency of the messaging and data transfer, for which Intel compensated by increasing bandwidth and clock speed of this interface. The expense of that is power and efficiency.

For an on-die interconnect to remain relevant, it needs to be flexible in bandwidth scaling, reduce latency, and remain energy efficient. With 28-core Xeon processors imminent, and new IO capabilities coming along with it, the time for the ring bus in this space is over.

Starting with the HEDT and Xeon products released this year, Intel will be using a new on-chip design called a mesh that Intel promises will offer higher bandwidth, lower latency, and improved power efficiency. As the name implies, the mesh architecture is one in which each node relays messages through the network between source and destination. Though I cannot share many of the details on performance characteristics just yet, Intel did share the following diagram.

intelmesh.png

As Intel indicates in its blog on the mesh announcements, this generic diagram “shows a representation of the mesh architecture where cores, on-chip cache banks, memory controllers, and I/O controllers are organized in rows and columns, with wires and switches connecting them at each intersection to allow for turns. By providing a more direct path than the prior ring architectures and many more pathways to eliminate bottlenecks, the mesh can operate at a lower frequency and voltage and can still deliver very high bandwidth and low latency. This results in improved performance and greater energy efficiency similar to a well-designed highway system that lets traffic flow at the optimal speed without congestion.”

The bi-directional mesh design allows a many-core design to offer lower node to node latency than the ring architecture could provide, and by adjusting the width of the interface, Intel can control bandwidth (and by relation frequency). Intel tells us that this can offer lower average latency without increasing power. Though it wasn’t specifically mentioned in this blog, the assumption is that because nothing is free, this has a slight die size cost to implement the more granular mesh network.

Using a mesh architecture offers a couple of capabilities and also requires a few changes to the cache design. By dividing up the IO interfaces (think multiple PCI Express banks, or memory channels), Intel can provide better average access times to each core by intelligently spacing the location of those modules. Intel will also be breaking up the LLC into different segments which will share a “stop” on the network with a processor core. Rather than the previous design of the ring bus where the entirety of the LLC was accessed through a single stop, the LLC will perform as a divided system. However, Intel assures us that performance variability is not a concern:

Negligible latency differences in accessing different cache banks allows software to treat the distributed cache banks as one large unified last level cache. As a result, application developers do not have to worry about variable latency in accessing different cache banks, nor do they need to optimize or recompile code to get a significant performance boosts out of their applications.

There is a lot to dissect when it comes to this new mesh architecture for Xeon Scalable and Core i9 processors, including its overall effect on the LLC cache performance and how it might affect system memory or PCI Express performance. In theory, the integration of a mesh network-style interface could drastically improve the average latency in all cases and increase maximum memory bandwidth by giving more cores access to the memory bus sooner. But, it is also possible this increases maximum latency in some fringe cases.

Further testing awaits for us to find out!

Source: Intel

Podcast #448 - Mesh Networking, Corsair ONE PRO, Windows 10 S, and Vega Specs

Subject: Editorial | May 4, 2017 - 10:15 AM |
Tagged: Windows 10 S, video, Vega, surface, Predator X27, podcast, ONE PRO, mesh, Intel, google wifi, eero, corsair, atom, Amplifi HD, acer

PC Perspective Podcast #448 - 05/04/17

Join us for mesh networking performance, Corsair ONE PRO, Microsoft / AMD / NVIDIA updates, 'leaked' Vega specs and more!

You can subscribe to us through iTunes and you can still access it directly through the RSS page HERE.

The URL for the podcast is: http://pcper.com/podcast - Share with your friends!

Hosts: Jeremy Hellstrom, Josh Walrath

Peanut Gallery: Alex Lustenberg

Program length: 54:15

Podcast topics of discussion:

  1. Week in Review:
  2. News items of interest:
  3. Hardware/Software Picks of the Week
  4. Closing/outro
 

Source:

Podcast #438 - Vulkan, Logitech G213, Ryzen Preorders, and more!

Subject: Editorial | February 23, 2017 - 12:16 PM |
Tagged: podcast, vulkan, ryzen, qualcomm, Qt, mesh, g213, eero, corsair, bulldog

PC Perspective Podcast #438 - 02/23/17

Join us for Vulkan one year later, Logitech G213 Keyboard, eero home mesh networking, Ryzen Pre Orders, and more!

You can subscribe to us through iTunes and you can still access it directly through the RSS page HERE.

The URL for the podcast is: http://pcper.com/podcast - Share with your friends!

Hosts: Allyn Malventano, Ken Addison, Josh Walrath, Jermey Hellstrom

Program length: 0:58:01

Podcast topics of discussion:
  1. Week in Review:
  2. News items of interest:
  3. Hardware/Software Picks of the Week
    1. Allyn: SS64.com - Nifty programmer's reference for scripting, web, db
    2. Ken: Dell refurbished XPS 13
  4. Closing/outro
 

Source:
Author:
Subject: Networking
Manufacturer: eero

Living the Mesh Life

Mesh networking is the current hot topic when it comes to Wi-Fi. Breaking from the trend of increasingly powerful standalone Wi-Fi routers that has dominated the home networking scene over the past few years, mesh networking solutions aim to provide wider and more even Wi-Fi coverage in your home or office through a system of multiple self-configuring and self-managing hotspots. In theory, this approach not only provides better wireless coverage overall, it also makes the setup and maintenance of a Wi-Fi network easier for novice and experienced users alike.

eero-1.jpg

Multiple companies have recently launched Wi-Fi mesh systems, including familiar names such as Google, Netgear, and Linksys. But this new approach to networking has also attracted newcomers, including San Francisco-based eero, one of the first companies to launch a consumer-targeted Wi-Fi mesh platform. eero loaned us their primary product, the 3-piece eero Home WiFi System, and we've spent a few weeks testing it as our home router.

eero-2.jpg

This review is the first part of a series of articles looking at Wi-Fi mesh systems, and it will focus on the capabilities and user experience of the eero Home WiFi System. Future articles will compare eero to other mesh platforms and traditional standalone routers, and look at comparative wireless performance and coverage.

Box Contents & Technical Specifications

As mentioned, we're looking at the 3-pack eero Home WiFi System (hereafter referred to simply as "eero"), a bundle that gives you everything you need to get your home or office up and running with a Wi-Fi mesh system. The box includes three eeros, three power adapters, and a 2-foot Ethernet cable.

eero-3.jpg

Each eero device is identical in terms of design and capability, measuring in at 4.75 inches wide, 4.75 inches deep, and 1.34 inches tall. They each feature two Gigabit Ethernet ports, a single USB 2.0 port (currently restricted to diagnostic use only), and are powered by two 2x2 MIMO Wi-Fi radios capable of supporting 802.11 a/b/g/n/ac. In addition, an eero network supports WPA2 Personal encryption, static IPs, manual DNS, IP reservations and port forwarding, and Universal Plug and Play (UPnP).

eero-4.jpg

Continue reading our early testing of the eero Home WiFi system!

CES 2017: ASUS Announces HiveSpot and HiveDot Mesh WiFi

Subject: Networking | January 4, 2017 - 07:40 PM |
Tagged: wifi, router, mesh, hivespot, hivedot, gigabit router, asus, 802.11ac

ASUS has just announced the HiveSpot and HiveDot Mesh WiFi systems, which both combine multiple access points into a single network. Any individual node could be configured as either a router or a repeater, but the system is designed around one acting as a router and the rest, repeaters. The main difference between the two models is the higher-end set, the HiveSpot, utilize an extra, 5 GHz band, running 867 megabit, that’s dedicated to communication between the access points.

ASUS-2017-ces HiveSpot  (3 HiveSpots).jpg

Because of this, the HiveSpot is listed as AC2134 while the HiveDot is AC1300, but devices that connect to this network will see two, 650 megabit bands in either case. What the HiveSpot will get you is higher performance (and maybe stability) should multiple devices be communicating with different nodes at the same time. With the HiveDot, the routers will be sharing the same bandwidth as the devices connecting to them.

ASUS wasn’t too clear about pricing in their press release, but CNet is reporting that they will be sold in bundles of three, which is the minimum for the mesh network. Three HiveSpot devices will carry an MSRP of $399 USD, while three HiveDots, $299. In other words, it will cost you $100 if you want the high-bandwidth, dedicated link between the nodes.

Coverage of CES 2017 is brought to you by NVIDIA!

PC Perspective's CES 2017 coverage is sponsored by NVIDIA.

Follow all of our coverage of the show at https://pcper.com/ces!

Source: ASUS

Google WiFi Bringing Wireless Mesh Networking to the Home

Subject: Networking | October 9, 2016 - 01:42 AM |
Tagged: wifi, onhub, mesh, google wifi, google, 802.11ac

Building on the company’s OnHub WiFi router program, the search giant will be offering up its own mesh WiFi network solution for home users later this year aptly named “Google WiFi.” Available in November for pre-order Google will offer single and triple packs of its puck-shaped smartphone controlled WiFi nodes.

Google WiFi node.png

Google WiFi is a new product that takes advantage of an old technology called mesh networking. While most home users rely on a single powerful access point to distribute the wireless signal throughout the home, mesh networks place nodes around the home in such a way that the WiFi networks overlap. Devices can connect to any node and transition between nodes automatically. The nodes communicate with each other wirelessly and connect end devices to the router and Internet by taking the best path (least number of hops and/or highest signal strengths). This model does have some disadvantages that are shared with WiFi repeater solutions in that as much as 50% (or worse!) of the bandwidth can be lost at each hop as the devices use wireless for both communicating with end devices and the backbone to the router. The advantage though is that you need only find a power outlet to set up the mesh node and there is no need to run Ethernet or deal with Powerline or MoCA setups.

Fortunately, it looks as though Google has mitigated the disadvantage by including two radios. The circular Google WiFi nodes (which measure 4.17” diagonally and 2.7” tall) pack a dual band 802.11ac WiFi chip that can operate at both 2.4 GHz and 5 GHz. Using the 5 GHz network for in room end devices (PCs, smartphones, game consoles, Rokus, et al) and the 2.4 GHz network to communicate with each other will help to eliminate a major bottleneck. There will likely still be some bandwidth lost, especially over multiple hops, due to interference, but it should be much less than 50% bandwidth loss.

Google WiFi Mesh.png

Each Google WiFi node features two Gigabit Ethernet ports that can be setup as LAN or WAN ports, Bluetooth, and an 802.11ac 2x2 WiFi radio with beamforming support. The nodes are powered by an unspecified quad core processor, 512MB DDR3L memory, and 4GB of eMMC flash storage. The nodes apparently draw as much as 15 watts.

Of course, being Google, the Google WiFi can be controlled using an Android or iOS app that allows the administrator to pause WiFi on a per-device basis (e.g. set time limits for children), monitor device bandwidth usage and prioritize traffic, and automatically apply firmware updates to mitigate security risks. Additionally, Google WiFi automatically configures each node to use the best channel and band to get the best performance that supports all devices.

The nodes currently come only in white and are constructed of plastic. There are blue LEDs around the middle of the puck shaped device. Google WiFi will be available for pre-order in November. A single node will cost $129 while a three pack will cost $299. Google is not first to the wireless mesh party but it looks like it will be competitively priced (the three pack is $200 cheaper than eero, for example).

This looks like it might be a simple to setup solution if you or your family are currently running a single access point that can’t quite cover the entire home. I don’t really see this as a product for enthusiasts, but it might be worth recommending to people that just want WiFi that works with little setup. I will have to wait for reviews to say for sure though.

What are your thoughts on Google WiFi?

Also read:

Source: Google
Author:
Manufacturer: NZXT

Introduction and Features

Introduction

tempest1.jpg

Courtesy of NZXT

NZXT added two mid-tower case offerings to its Tempest series today that feature custom solutions for dual-radiator watercooling systems and dual "touch-powered" 120mm front fans with removable filters. The Tempest 410 and Tempest 410 Elite are available for $79.99 and $89.99. Today, we are kicking the tires on the Tempest 410 Elite to ensure enthusiasts and overclockers get the most bang for their buck for their next mid-tower case upgrade.

tempest2.jpg

Courtesy of NZXT

Both Tempest 410 and Tempest 410 Elite cases incorporate a honeycomb mesh design to the front, top, and back panels and advanced cable management systems for concealing loose wires and power cords. The Tempest 410 Elite ups the ante with an acrylic side panel to see all the hardware inside the chassis.

Continue reading our review of the NZXT Tempest 410 Elite Mid-Tower Case!!