Hitachi gung-ho for Nehalem EP blades

Mum on Itanium

Hitachi Server Group today rolled out its BladeSymphony blade servers based on Intel's Nehalem EP Xeon 5500 quad-core processors.

The Server Group, which sells servers in North America, has updated its existing entry-level BladeSymphony 320 machines to adopt the Nehalem chips, and debuted a new high-end blade server, the BladeSymphony 2000, aimed at supporting larger workloads.

Hitachi hasn't had a huge presence in North America since Hitachi Data Systems sold bi-polar mainframes that were better than alternatives supplied by IBM a decade and a half ago - HDS is now relegated to selling storage outside of its home Japanese market. However, Hitachi has been ramping up its presence in North America over the past few years, doing so solely with its BladeSymphony blade servers.

The BladeSymphony servers debuted back in July 2005 when Intel upgraded its single-core Madison Itanium processors. Back then, the BladeSymphony 1000 was an Itanium-only box with up to eight blade slots in its 10U chassis.

The BladeSymphony setup was a bit different from typical blade servers in that it supported Itanium and had its own hardware partitioning and virtualization technology, Virtage, that not only allowed for blades to be carved up to support multiple workloads, but also allowed the eight two-socket Itanium blades to be converted into a 16-way SMP server with 256GB of main memory.

As Intel doubled up the Itanium cores with the Montecito 9000 series and Montvale 9100 series Itaniums, the chipset used in the BladeSymphony box allowed 16-way SMPs to be made out of four blades. In November 2006, Hitachi started selling this product in North America, and also created two-socket Xeon blades for the chassis, which were based on the Xeon DP chips and therefore didn't have the funky SMP magic.

In April 2007, Hitachi kicked out the BladeSymphony 320 chassis, a 6U box that supported half-height blades and sported ten blades. That fall, Hitachi ported the Virtage hardware-partitioning microcode (which runs on the service processor in the chassis) to the BladeSymphony 320, which was also given a 110-volt option for office environments. As Intel rolled out new Xeon DP processors, Hitachi put them into the 320 chassis where appropriate.

That brings us to today and the new BladeSymphony 2000 10U chassis with room for ten full-height blades. If that doesn't sound as dense as some of the alternative blade gear out there from IBM, Hewlett-Packard, Dell, and Fujitsu, it's because it isn't.

But Steve Campbell, VP of marketing and solutions at Hitachi Server Group, says that the company is not interested in skimping even a little bit on CPU, memory, or I/O capacity just for the sake of density. And that, he says, is because customers don't want to make those sacrifices.

The blades used in the BladeSymphony 2000 chassis support only the Nehalem EP chips and offer 144GB of main memory (using 8GB DDR3 memory modules) and up to 162Gb/sec of I/O bandwidth per blade.

The BladeSymphony has what Hitachi calls a hybrid I/O subsystem, which uses PCI-Express 2.0 I/O slots paired with integrated switches, allowing the Virtage virtualization environment to virtualize I/O across blades while also allowing specific PCI-Express peripherals to be bound to specific blades.

Campbell says that the new Nehalem EP blades have about 50 per cent better performance than prior two-socket Xeon blades, four times the main memory, and seven times the I/O bandwidth.

Hitachi makes its own Ethernet switches for the BladeSymphony lineup and resells Fibre Channel switches from Cisco Systems and Brocade Communications to link blades to storage. Campbell says that Hitachi is closely watching developments in converged enhanced Ethernet and Fibre Channel over Ethernet, but adds that as far as Hitachi is concerned, FCoE is not yet ready for prime time until the standard settles down and Ethernet can be demonstrated to support storage traffic without dropping bits.

Looking ahead, it's likely that Hitachi will support eight-core Nehalem EX Xeon 7500 processors on blades that slide into the BladeSymphony 2000 chassis. The Virtage clustering and partitioning technology is not limited to 16-way SMP, so Hitachi could, in theory, make a pretty powerful SMP box out of a 10-socket blade chassis using Nehalem EXs.

As for Hitachi Server Group's plans for the quad-core Tukwila Itanium chips - which are perhaps coming soon - Campbell didn't let on much. He did, however, say that the market in North America seems to be more interested in Nehalem blades and the performance they will provide, quickly adding that the BladeSymphony 1000 machines are still being sold and are being adopted by customers to run heavier workloads.

Because of the QuickPath Interconnect ports on the Nehalem chips and their inherent support of NUMA clustering, Virtage SMP clustering can be tweaked to work on future Nehalem EX ports much as it did for the earlier Itanium chips. The Virtage clustering doesn't have an inherent 16-core limit (which apparently has more to do with the Itanium architecture's bandwidth on the front side bus), so it would be interesting to see a big bad SMP machine using the Nehalem EXs. Once you get beyond 32 cores, the performance benefits drop off pretty fast with SMP - but with five double-wide four-socket Nehalem EX blades, you're talking about 160 cores. Even half that would be pretty impressive.

This is all theory, of course. Campbell wouldn't say exactly what Hitachi has cooking for Nehalem EX, but he did say that the company has no interest in using Advanced Micro Devices' Opteron processors in its blade machines - as has been true for the past four years.

The existing BladeSymphony 320 chassis has also been given a bunch of Nehalem EP blades, including a standard two-socket blade with two hot-swap SAS disks, another blade designed for linking out to SANs that don't have local storage, and a third that has two Nehalem chips and room for six 300GB SAS drives for setups that need a lot of local storage. Hitachi is also readying a PCI blade for later this year - and it's not at all clear what that is because the feeds and speeds of the new BladeSymphony 320 blades are not yet out.

Both the BladeSymphony 2000 and updated BladeSymphony 320 will be available on June 1st. Entry pricing for a base BladeSymphony 2000 - including the chassis, one blade server with a reasonable configuration, and an Ethernet switch - will run to around $10,000. A base Nehalem EP blade for the BladeSymphony 320 will cost around $3,500, including the chassis. ®

Similar topics

Other stories you might like

  • GPL legal battle: Vizio told by judge it will have to answer breach-of-contract claims
    Fine-print crucially deemed contractual agreement as well as copyright license in smartTV source-code case

    The Software Freedom Conservancy (SFC) has won a significant legal victory in its ongoing effort to force Vizio to publish the source code of its SmartCast TV software, which is said to contain GPLv2 and LGPLv2.1 copyleft-licensed components.

    SFC sued Vizio, claiming it was in breach of contract by failing to obey the terms of the GPLv2 and LGPLv2.1 licenses that require source code to be made public when certain conditions are met, and sought declaratory relief on behalf of Vizio TV owners. SFC wanted its breach-of-contract arguments to be heard by the Orange County Superior Court in California, though Vizio kicked the matter up to the district court level in central California where it hoped to avoid the contract issue and defend its corner using just federal copyright law.

    On Friday, Federal District Judge Josephine Staton sided with SFC and granted its motion to send its lawsuit back to superior court. To do so, Judge Staton had to decide whether or not the federal Copyright Act preempted the SFC's breach-of-contract allegations; in the end, she decided it didn't.

    Continue reading
  • US brings first-of-its-kind criminal charges of Bitcoin-based sanctions-busting
    Citizen allegedly moved $10m-plus in BTC into banned nation

    US prosecutors have accused an American citizen of illegally funneling more than $10 million in Bitcoin into an economically sanctioned country.

    It's said the resulting criminal charges of sanctions busting through the use of cryptocurrency are the first of their kind to be brought in the US.

    Under the United States' International Emergency Economic Powers Act (IEEA), it is illegal for a citizen or institution within the US to transfer funds, directly or indirectly, to a sanctioned country, such as Iran, Cuba, North Korea, or Russia. If there is evidence the IEEA was willfully violated, a criminal case should follow. If an individual or financial exchange was unwittingly involved in evading sanctions, they may be subject to civil action. 

    Continue reading
  • Meta hires network chip guru from Intel: What does this mean for future silicon?
    Why be a customer when you can develop your own custom semiconductors

    Analysis Here's something that should raise eyebrows in the datacenter world: Facebook parent company Meta has hired a veteran networking chip engineer from Intel to lead silicon design efforts in the internet giant's infrastructure hardware engineering group.

    Jon Dama started as director of silicon in May for Meta's infrastructure hardware group, a role that has him "responsible for several design teams innovating the datacenter for scale," according to his LinkedIn profile. In a blurb, Dama indicated that a team is already in place at Meta, and he hopes to "scale the next several doublings of data processing" with them.

    Though we couldn't confirm it, we think it's likely that Dama is reporting to Alexis Bjorlin, Meta's vice president of infrastructure hardware who previously worked with Dama when she was general manager of Intel's Connectivity group before serving a two-year stint at Broadcom.

    Continue reading

Biting the hand that feeds IT © 1998–2022