AMD Hints To Hyper-Threading In 2012

April 23, 2009

Ahead in many fronts, AMD is still playing catch up to Intel on hyper-threading.

Some of you have raised the issue on why AMD has not adopted hyper-threading to increase performance of its chips in highly threaded environment such as virtualization.

Recently a TechPulse360 reader commented on why hyper-threading made sense:

  1. Hyperthreading gives tremendous boost in certain applications. 10-15% is towards the lower end of the scale;
  2. All modern CPU architectures (except AMD) support multiple threads per core. Look at POWER from IBM, T1/T2/Rock from Sun Micro, Nehalem/Atom/Larrabee from Intel …;
  3. A proper implementation of simultaneous multithreading (SMT, and what Intel calls Hyper-Threading) requires few additional resources. There is no such thing as a “Normal Pipeline” and a “Hyperthreaded pipeline” – all the functional units that form the bulk of the pipleline are unchanged, certain resources are shared, while certain other resources (like ISA registers) are duplicated.

This week an AMD engineer confided to me that not having hyper-threading available made Opteron look slower than Intel’s low-end chips. The engineer also said that people at AMD have now admitted that not having hyper-threading was the wrong technical choice.

So here’s what Pat Patla, AMD’s server boss had to say when I asked him about hyper-threading during my visit at AMD’s Sunnyvale, Calif.- headquarters this week:

“If you look at our future roadmap and what we’re showing for adressing the threaded market, we believe it is best addressed at full core count this time. And you saw our 2010 time frame when we are talking about 12 cores per CPU and in 2011 with 16 cores per CPU. So we think we are pretty well covered in the 48 to 64 threads environment for the next couple years and we’ll see what 2012 and 2013 brings.”

It sounds to me that AMD’s 2012 chips are going to have hyper-threading!

Here’s a video excerpt of Pat Patla answers on the hyper-threading question:


Intel Core i7 Is Fastest High-End Consumer Processor, Cuts “Northbridge” To Nowhere; But AMD Phenom II Will Be Fastest Chip For Mainstream PCs In 2009

November 16, 2008
Nathan Brookwood, analyst at Insight64

Nathan Brookwood, analyst at Insight64

With the launch this week-end of Core i7, Intel released the world’s fastest high-end consumer chip. But as Insight64 analyst, Nathan Brookwood pointed out in a conversation today, AMD’s next generation desktop chip, the Phenom II, will still be the top performer in the mainstream/value PC market.

“Intel will not release a mainstream version of Core i7 before the end of next year. During that period, AMD Phenom II will be the fastest chip in the $1,200 or less PC category beating Intel’s old generation of quad-core chips. But this is also a tacit admission on AMD’s part that they can’t play in the high-end space anymore,” said Brookwood.

Core i7 code named Nehalem will first be available for high-end desktop PCs, workstations and uni-processor servers tomorrow, then for 2-way servers in Q109 and 4-way servers for second half 2009.

Nehalem cuts the “Northbridge” to nowhere

For Brookwood, Core i7 is Intel’s first radical architecture since it began making processors in the early 1970s.

“Nehalem as a family of products represents a major change in way Intel products will work with higher level of performance for the next several years.

It’s also the beginning of the end for Intel’s northbridge system architecture which has characterized Intel’s system since the beginning of Intel and microprocessors. Intel X86 microprocessors always relied on logic outside of the CPU itself to handle memory acceses and interprocessor kind of operations. And now Intel is making the same kind of changes than AMD made back in 2003, 6 years ago with the onboard memory controller.

As this technology ripples through the rest of Intel’s product line, with the entire Intel product line rolling over Nehalem by early 2010, we will see the northbridge to nowhere going away!,” adds Brookwood.

Hyper-Threading is back!

But perhaps the single most difference between Intel and AMD architectures is the support of hyper-threading in the Core i7 processor family, a technology that was sort on a “sabbatical” since the Pentium 4 faded away. A hyper-threaded processor is treated by the operating system as two processors instead of one.

“Hyperthreading provides 15% to 20% more performance with relatively little increase in silicon size. And these days, it’s hard to find 20% more performance. So when you know how to do that you certainly want to pursue those kind of opportunities,” explains Brookwood.

According to Brookwood, Intel left out hyper-threading in its last generation Core architecture in order to focus on other priorities knowing that it will still have substantial performance with Core, even without the hyper-threading technology.

But all in all, with Nehalem, AMD and Intel will now compete with chips having a nearly identical architecture. Which makes Brookwood believe that AMD could not leave Intel in the dust and vice-versa.


Follow

Get every new post delivered to your Inbox.

Join 32 other followers