Chipzilla unveils six-core 'Dunnington' Xeons

Edges AMD on big X64 iron


With virtualization and server consolidation bolstering sales of larger - and more profitable - servers, Intel wants to keep the competitive pressure on its rival in the X64 arena, Advanced Micro Devices, while also giving its server partners the chance to go after big RISC/Unix boxes in the same power class.

With the "Dunnington" six-core Xeon 7400 processors announced today, the X64 platform moves deeper into RISC territory and puts some water between Intel's offerings and those from AMD, which top out at four cores with the "Barcelona" Opteron 8300 series.

The Dunnington chip is socket-compatible with the existing "Tigerton" Xeon 7300 MP processor, and both are designed to be used in machines with four or more processor sockets. The Dunnington processor is based on the "Penryn" core that launched nearly a year ago, and the cores are implemented in a 45 nanometer, Hi-K process that allows the top-end chips to deliver as much as 50 percent more performance than the comparable 7300s while also consuming 10 percent less power.

The Dunnington Xeons use a 1.07 GHz front side bus and have 96 KB of L1 cache per core and 3 MB of L2 cache per core pair. The chip has 16 MB of shared L3 cache on the die - the first time the Xeon architecture has had L3 cache on the chip.

Like the other Xeons since the "Woodcrest" dual-core Xeons from several years ago, the Dunningtons have support for a virtualization feature called VT FlexMigration, which ensures that virtual machine hypervisors and their teleportation features (which allow VMs to be moved around a network of machines) are both forward and backward compatible.

The chip does not, like its Tigerton predecessor, have HyperThreading, which presents two virtual instruction streams to an operating system for every real one on the chip. HyperThreading tends to boost performance by about 20 to 30 percent in the X64 architecture, but it also eats up die space.

The Dunnington chip has 1.9 billion transistors, making it one of the largest chips Intel has ever made in terms of transistor count - second only perhaps to the "Tukwila" quad-core Itanium, which is expected to come in at just over 2 billion transistors.

Depending on how little money you have, Intel turns off cores and cache until you meet your budget. Well, sort of. The top-end X7640 has all six cores activated, runs at 2.66 GHz, and has 16 MB of L3 cache; it has a 130 watt thermal design point (TDP) and costs $2,729 in 1,000-unit quantities. (All Intel processor prices are per chip if you buy a 1,000 chip tray all at once.) The practical Dunnington for most users, given budget and thermal constraints, is probably the E7450, which also has all six cores running, but only 12 MB of L3 cache and a clock speed of only 2.4 GHz, which drops the thermals down to 90 watts and the price down to $2,301.

The only other six-core Dunnington is a low-voltage variant, the L7455, which runs at 2.13 GHz and has 12 MB of L3 cache and sports a 65 watt TDP. It costs $2,729, just like the faster 130 watt part, which means if you want the best performance or the best thermals, you pay either way.

The E7440 is the fastest quad-core variant of Dunnington, which runs at 2.4 GHz and has 16 MB of L3 cache at a price of $1,980, and cutting the cache back to 12 MB and reducing clock speed on a quad to 2.13 GHz drops the price further on the E7430 to $1,391. The E7420 runs at 2.13 GHz as well, but has only 8 MB of cache and a price of $1,177.

All three of these four-core Dunningtons have a 90 watt TDP. The remaining Dunnington is a low-voltage L7445 part, which has four cores active, running at 2.13 GHz and accessing a 12 MB L3 cache. It costs $1,980.

Tom Kilroy, general manager of Intel's Digital Enterprise Group, said at the Dunnington launch in San Francisco that the number of virtualized servers out there in the data centers of the world has grown by 50 percent in the past year, and that server consolidation is a big driver behind sales of the kinds of big boxes that chips such as the Xeon 7400 series engender. Of course, needing to run massive ERP suites and their related databases, or huge email servers, also drives sales of big iron.

According to benchmark data presented across a number of different workloads by Kilroy, the 7400s in four-socket configurations using six cores best the 7300s using quad cores by between 14 and 48 percent. The weakest performance improvement is on e-commerce workloads, with the average being somewhere close to 35 percent. The best performance Intel showed for the Dunnington chip compared to its predecessor was for database processing as it relates to the relatively new TPC-E benchmark test. On the TPC-C test, the performance boost was only 43 percent.

Kilroy said that in the past year, since Intel has started shipping 45 nanometer processors, the company has shipped close to 50 million units using the chip making technology. ®

Similar topics

Broader topics


Other stories you might like

  • AMD claims its GPUs beat Nvidia on performance per dollar
    * Terms, conditions, hardware specs and software may vary – a lot

    As a slowdown in PC sales brings down prices for graphics cards, AMD is hoping to win over the market's remaining buyers with a bold, new claim that its latest Radeon cards provide better performance for the dollar than Nvidia's most recent GeForce cards.

    In an image tweeted Monday by AMD's top gaming executive, the chip designer claims its lineup of Radeon RX 6000 cards provide better performance per dollar than competing ones from Nvidia, with all but two of the ten cards listed offering advantages in the double-digit percentages. AMD also claims to provide better performance for the power required by each card in all but two of the cards.

    Continue reading
  • Google opens the pod doors on Bay View campus
    A futuristic design won't make people want to come back – just ask Apple

    After nearly a decade of planning and five years of construction, Google is cutting the ribbon on its Bay View campus, the first that Google itself designed.

    The Bay View campus in Mountain View – slated to open this week – consists of two office buildings (one of which, Charleston East, is still under construction), 20 acres of open space, a 1,000-person event center and 240 short-term accommodations for Google employees. The search giant said the buildings at Bay View total 1.1 million square feet. For reference, that's less than half the size of Apple's spaceship. 

    The roofs on the two main buildings, which look like pavilions roofed in sails, were designed that way for a purpose: They're a network of 90,000 scale-like solar panels nicknamed "dragonscales" for their layout and shimmer. By scaling the tiles, Google said the design minimises damage from wind, rain and snow, and the sloped pavilion-like roof improves solar capture by adding additional curves in the roof. 

    Continue reading
  • Pentester pops open Tesla Model 3 using low-cost Bluetooth module
    Anything that uses proximity-based BLE is vulnerable, claim researchers

    Tesla Model 3 and Y owners, beware: the passive entry feature on your vehicle could potentially be hoodwinked by a relay attack, leading to the theft of the flash motor.

    Discovered and demonstrated by researchers at NCC Group, the technique involves relaying the Bluetooth Low Energy (BLE) signals from a smartphone that has been paired with a Tesla back to the vehicle. Far from simply unlocking the door, this hack lets a miscreant start the car and drive away, too.

    Essentially, what happens is this: the paired smartphone should be physically close by the Tesla to unlock it. NCC's technique involves one gadget near the paired phone, and another gadget near the car. The phone-side gadget relays signals from the phone to the car-side gadget, which forwards them to the vehicle to unlock and start it. This shouldn't normally happen because the phone and car are so far apart. The car has a defense mechanism – based on measuring transmission latency to detect that a paired device is too far away – that ideally prevents relayed signals from working, though this can be defeated by simply cutting the latency of the relay process.

    Continue reading

Biting the hand that feeds IT © 1998–2022