Intel Discusses The Future Of Many Integrated Core (MIC) Products

Intel Discusses The Future Of Many Integrated Core (MIC) Products

Larrabee, Intel's once-vaunted, next-generation graphics card died years ago, but the CPU technology behind the would-be graphics card has lived on. Intel discussed the future of MIC/Knight's Corner today. After Larrabee was officially canceled, Intel repurposed the design and seeded development kits to appropriate market segments. MIC cards won't start shipping until the 22nm Knight's Corner chip is launched, but even the Knight's Ferry prototypes offer tantalizing hints at what future performance might be resemble.

Like Larrabee, Knight's Corner (and future MIC products in general) utilize a CPU based on Intel's original Pentium architecture (P54C). Modifications include complete cache coherency, x86-64 compatibility, and 512-bit vector support capable of performing 16 single-precision floating point operations simultaneously.

One advantage Intel is talking up in a big way is the fact that existing x86 applications will be relatively easy to port to MIC processors. The exact amount of necessary modification will depend on what the program does and how easy it is to parallelize the workload across the 50 chips Knight's Corner will offer. The company claims that "With greater reuse of parallel processor code, software companies and IT departments benefit from creating and maintaining a single code base binary and not having to retrain developers on proprietary programming models associated with accelerators."


The advantages of Knight's Crossing, from Intel's point of view.

That's true to a point but it's not universally applicable. Knight's Corner is almost certainly 9-12 months away, if not longer. That's more than enough time for NV and AMD to make their own move to 28nm and further evolve their own GPGPU architectures. This is more of a potential issue for Nvidia than AMD--Team Green doesn't currently have a Tesla-equivalent GPGPU product line.

x86 compatibility may give Intel an edge, but it's far from the only important factor. Knight's Corner will utilize Intel's tri-gate technology, but will rely, at least in part, on proprietary Intel development tools and the OpenMP software standard. According to SGI's CTO, Eng Lim Goh, Knight's Corner is intriguingly easy to program. Performance, however, varies. Speaking to the EETimes, Goh noted:

[P]erformance ranges dramatically based on applications from orders of magnitude improvements to incremental improvements" using the current chips [Knight's Ferry] that support only single-precision floating point operations.

You end up with two kinds of customers, one highly satisfied with [AMD and Nvidia graphics] accelerators because despite the tedious porting process, their results are very good. Others feel their time spent on porting [their apps to AMD and Nvidia chips] doesn’t justify the performance and there is a huge part of this second group for whom MIC is useful—and ultimately some of the first group may want MIC, too.
The one good sign (from NV's perspective) is that while Intel mentioned 'exaflop computing' multiple times in its presentation, the company refrained from publishing any benchmark results between Knight's Ferry/Knight's Corner products and any of NV's Tesla GPUs. That could mean that Intel isn't completely confident in how the two solutions stack up against each other.

Alternatively, Santa Clara may not want to tip its hand too early. Larrabee may be dead and gone, but its legacy could still mean trouble for Nvidia.
 
0
+ -

Some amazing Knight's Corner technology combined with Intel's monopolistic business practices will be an unbeatable combination - as if the ability to bribe Dell 6-billion dollars not to use AMD chips wasn't already enough.

Exciting times!

0
+ -

50 cores on one dies sounds unbelievable though I know there are prototypes that have 100's of cores. While this is impressive I think some of the technology that uses multiple types of cores efficiently is a bit more promising as far as performance is concerned. CPUs are good at certain types of calculations and GPUs are better at other types.

0
+ -

I agree Draco. I can see AMD's Fusion tech having less cpu capability then this, but much more rounded performance and a massive advantage in GPU compatible actions.

0
+ -

This is almost like the 100 core CPU that Intel showed off. Except this is with fewer cores and it's going to be in the consumers hands.

Imagine the possibilities that one could do with 30 cores. Seriously, that's some computer processing power right there.

0
+ -

This isn't going to be in consumers' hands unless you're working at a university or HPC lab. Unless you've time and resources to do a great deal of recompiling, the best native performance on a Knight's Corner would offer 2-4 cores at most.

Anyone who thinks 2-4 Pentium-class cores are going to offer something superior to Sandy Bridge in terms of general performance is incorrect.

0
+ -

That was very interesting to read about future Intel Processor. Time will tell when it come out as preview or something will show up in future.

0
+ -

This processor kinda weirds me out. Is intel making a graphics processor that can only do compute?

0
+ -

Intel still has no answer for Fusion. For pure power computing and number crunching and such....absolutely Intel has the edge.

Unfortunately for them (Intel) AMD had the complete package going forward as far as a solid CPU and a solid GPU to build upon.

Instead of scrapping the whole idea and man hours that went into this project knows as Larrabee Intel will find a use for it in some niche market.

0
+ -

Thunderbird,

Larrabee may not have come to market as Intel initially intended, but the concept of putting many simple cores on a single die is extremely pertinent. The point of the story isn't that Intel took a failed GPU and shoehorned it into a niche market. It's more accurate to say that Intel saw a market for devices that featured a relatively large number of x86 cores. Said market encompasses everything from small handheld devices to large-scale HPC implementations.

You could read this as Intel agreeing with Nvidia that there's a market for highly parallel CPUs. The two companies are approaching the concept very differently,but they share a core belief in the viability of their chosen vectors.

Edit: Additionally, the GPU inside the next-generation 22nm Ivy Bridge Intel will launch at the end of this year or the beginning of 2012, will be DX11-compatible and fully programmable. We don't know much about its specific capabilities but just as Sandy Bridge was a huge leap over integrated 45nm for Nehalem, IB will leap again over SB.

0
+ -

@ joel

You are correct but Intel did indeed intend for Larabee to compete with Fusion. 

Kudos to Intel for finding a market for the many X86 core package that's what good companies do they market themselves.

I myself am using an Intel processor and agree that Ivy Bridge will offer better performance than Sandy Bridge.

Login or Register to Comment
Post a Comment
Username:   Password: