Posts tagged with Istanbul
Recap of early “Istanbul” Reviews
Posted by Phil Hughes in 8:51 AM
(4 votes, average: 4.00 out of 5)
Loading ...On June 1 we launched our Six-Core AMD Opteron™ processor, formerly code-named “Istanbul”. Here are some of the early highlights from some of the industry’s most-respected hardware reviewers.
AnandTech: AMD’s Six-Core Opteron 2435
“It is out of the scope of this article, but it is clear that even if the CPUs cost the same, the AMD based server will be less costly.”
“Our power consumption data is preliminary. We really have to double-check all the power data. Very roughly, we find that the Opteron 2435 machine consumes about 35-45W less than the Xeon X5570. On a total of slightly more than 300W, that is about 10 to 15%. Idle power seems to be slightly in favor of the Xeon “Nehalem”.
“There are two types of applications where we feel that the AMD six-core deserves your attention: decision support databases and virtualization. . . . Of course, performance-critical applications are by definition more demanding when it comes to processing power. That is exactly what vApus Mark I measures: how well do performance-critical applications perform when they are virtualized? This is a relatively “new” market where the AMD 2435 shines. The new Opteron 2435 at 2.6 GHz was a pleasant surprise on vApus Mark I: it keeps up with more expensive Xeons on ESX 3.5 update 4 while consuming less, and offers a competitive performance/watt and performance/price ratio on vSphere 4. The six-core Opteron is about 11 to 30% slower on vSphere 4 than the 2.93 GHz Xeon X5570 but the overall cost of the Istanbul platform is significantly lower (DDR-2 versus DDR-3) and the 2.6 GHz 2435 consumes less power in a virtualized environment (*). On the condition that you optimize your hypervisor well to take advantage of the six cores (cell size is for example one critical optimization), we feel that the six-core Opteron is a worthy opponent for the Xeon “Nehalem” in this market. We tested only the 2435 versus the X55xx series.”
(*) Virtualized servers do not run idle very often
Tech Report: “AMD’s ‘Istanbul’ six-core Opteron processors / We crack open a six-pack”
AMD Executing Early: “AMD had originally planned to introduce this product in the October time frame, but the first spin of Istanbul silicon came back solid, so the firm pulled the launch forward into June.
“The most notable change, though, is probably the addition of a feature AMD calls HT Assist. HT Assist is essentially a probe filter intended to reduce the overhead required for the synchronization of cached data across CPUs in multiple sockets. HT Assist reserves space in each processor’s L3 cache, in which it stores an index of where that CPU’s cache lines are being used system-wide. The CPU then becomes “host” of the cache lines stored in its directory. If any CPU needs an update about a particular cache line, it will often know which CPU is the correct host to probe for that information. This reduction in probe traffic can yield big gains in available system bandwidth, as we reported when we saw AMD demo a 4P system whose Stream bandwidth increased from roughly 25GB/s to 42GB/s with the addition of Istanbul processors with HT Assist.”
“Yet Istanbul should be a clear improvement over Shanghai for many workstation-class workloads and most server-class workloads—i.e., those that are essentially parallel and widely multithreaded. The Opteron 2435 manages to deliver this higher performance not just within the same power envelopes, but quite empirically with almost the exact same measured power consumption as the Opteron 2389.”
“This combination yields a nice increase in power efficiency, which was enough to put our Istanbul-based test system in the same territory as our Xeon X5550 system. The competition between the two was remarkably close in SPECpower_ssj, and the Istanbul system required notably less energy to render the Cinema 4D sample scene in Cinebench. So despite that fact that Intel leads in outright performance, the Opteron 2435 is entirely competitive on the power-efficiency front, with lower peak power draw, to boot. Those who evaluate systems strictly on this basis would do well to keep Opterons in the mix.”
“And if you have existing, compatible Socket F servers, the Istanbul Opterons should be an excellent drop-in upgrade. They’re a no-brainer, really, when one considers energy costs and per-socket/per-server software licensing fees.”
ZDNet: “AMD Istanbul: Field Upgrade Only If Your Hands are Nimble”
Unlike Intel’s Nehalem, which requires entirely new and unproven mainboards, systems, chipsets and sockets, the “Istanbul” is based on existing, mature AMD chipsets and uses the existing 1207-pin Socket F, allowing current Quad-Core Opteron systems to be field upgraded rather than requiring entirely new server purchases. . . . . The 24xx and 84xx series CPUs are actually compatible with systems going back four Opteron quad-core chip generations, ever since the 2000 and 8000 series were introduced in August of 2006. . . . . AMD also was able to engineer and manufacture the Istanbul during a period in which it also spun off its entire manufacturing capacity as a separate company, GlobalFoundaries, so the fact they were able to get this product out in record time (six months ahead of schedule, according to the company) to compete with Intel’s Nehalem is a significant achievement.”
Phil Hughes is a senior public relations manager at AMD. His postings are his own opinions and may not represent AMD’s positions, strategies or opinions. Links to third party sites are provided for convenience and unless explicitly stated, AMD is not responsible for the contents of such linked sites and no endorsement is implied.
As the See-Saw Cycles: Congrats to Intel, with a few questions as well
Posted by John Taylor in 1:11 PM
About five months ago, AMD launched a server processor codenamed “Shanghai,” about which Anandtech said:
“It is clear that the latest AMD Opteron is in the lead. We are really at the pivotal moment in time. No matter how good the current Xeon “Harpertown” and “Dunnington” architectures are, they lose too many battles due to the platform they are running on. The FSB architecture is singing its swan song.
Today, it’s Intel’s turn in this see-saw dynamic as it launches its new architecture for 2P servers and workstations. The follow-on Anandtech review this week shows strong initial benchmark performance for the highest-end “Nehalem” platforms. Let’s congratulate Intel on the performance of its all-new processors and the attending all-new OEM platforms. Competition is good for the industry. With “Istanbul” on deck for 2H09, and AMD’s all new “G34” platforms following thereafter, this competition is far from over. Here’s a little more on Istanbul:
How are the news and social media telling Intel’s big story thus far? Some already remarked on similarities between Xeon 5500 Series processors and AMD Opteron innovations[1], while we expect the most reactionary and binary (declaring AMD or Intel sucks/rules based on mostly academic benchmarks) tech press to write misguided AMD Opteron obits today.
Others will write-up even the boldest Intel hyperbole like a speckled trout takes a shrimp. Speaking of running with what Intel gives you, it’s become a source of amusement in the AMD Communications team: What new way will Intel spin its beloved “The biggest X Intel innovation since Y Intel innovation?”
· For this launch, “Pat Gelsinger…touted the new chips…as the most significant processor launch since the Pentium Pro was released in 1995,” as quoted in eWeek
· However, Core 2 Duo was “One of the biggest new microprocessors we’ve done in many years--at least a decade--and many people say the biggest one since the Pentium,” said David Tuhy, general manager of desktop products division, in an interview with eWEEK.
· With High-K Metal Gate and 45nm, “Company researchers said the advance represented the most significant change in the materials used to manufacture silicon chips since Intel pioneered the modern integrated-circuit transistor more than four decades ago,” according to the New York Times.
To my ears, the Intel “biggest X since Y” tactic is now “the most-worn-out-yet-mysteriously-effective marketing tactic since the infomercial.”
While I may not get my way on news and social media tiring of an overdone marketing tactic, perhaps as the week progresses we’ll see answers to Xeon 5500 questions like these:
1. Current platform pricing shows that Xeon 5500 systems represent a significant price premium of up to 40 to 50% over equivalent AMD Opteron processor-based systems. Will customers see sufficient value at this price premium in this economy?
2. Are all the new features available across the entire Nehalem product stack? For example, do they have to reduce memory DIMM performance, lower their Quickpath speeds in order to achieve lower TDPs? Do all the models feature HyperThreading or Turbo Mode?
3. We’re hearing much emphasis with this launch on idle power. Will “Nehalem” EP platforms consume more power at the wall versus similar AMD Opteron processor-based systems at load, as in virtualization environments?
4. Intel termed 4P server configurations “for mission critical applications” and “with Virtualization [Intel has] seen four socket and above have more strength because there’s a larger desire to move to a consolidated environment because of the operational benefits that come with it.” When will Intel have 4P and higher solution based on its Nehalem architecture available and shipping in OEM systems?
5. Nehalem EP is only available with DDR3 memory, which draws more power and comes at a price premium. As servers are typically configured with massive amounts of memory, will Intel need to compensate for these two considerations?
Maybe we can get a little help from Jon Stewart in asking the questions that get missed?
John Taylor is Director, Global Platform & Product Communications at AMD. His postings are his own opinions and may not represent AMD’s positions, strategies or opinions. Links to third party sites are provided for convenience and unless explicitly stated, AMD is not responsible for the contents of such linked sites and no endorsement is implied.
[1] Excerpt from subscription-only Wall Street Journal story:
One reason for the speed of Nehalem — a family that includes models for high-end PCs introduced in November — is that it connects directly to memory chips rather than going through an external component called a memory controller. AMD has offered that technology on its Opteron line since 2003.
(4.90 out of 5)