Featured Articles

IHS teardown reveals Galaxy S5 BOM

IHS teardown reveals Galaxy S5 BOM

Research firm IHS got hold of Samsung’s new flagship smartphone and took it apart to the last bolt to figure out…

More...
Galaxy S5, HTC One M8 available selling well

Galaxy S5, HTC One M8 available selling well

Samsung’s Galaxy S5 has finally gone on sale and it can be yours for €699, which is quite a lot of…

More...
Intel lists Haswell refresh parts

Intel lists Haswell refresh parts

Intel has added a load of Haswell refresh parts to its official price list and there really aren’t any surprises to…

More...
Respawn confirms Titanfall DLC for May

Respawn confirms Titanfall DLC for May

During his appearance at PAX East panel and confirmed on Twitter, Titanfall developer Respawn confirmed that the first DLC pack for…

More...
KFA2 GTX 780 Ti Hall Of Fame reviewed

KFA2 GTX 780 Ti Hall Of Fame reviewed

KFA2 gained a lot of overclocking experience with the GTX 780 Hall of Fame (HOF), which we had a chance to…

More...
Frontpage Slideshow | Copyright © 2006-2010 orks, a business unit of Nuevvo Webware Ltd.
Wednesday, 11 August 2010 11:11

Intel unveils Knights Ferry HPC Co-Processor specifications

Written by Jon Worrel
intel_logo_new  intel_22nm_logo

32 cores, 128 threads, 8MB shared L2 cache, 1.20GHz
Earlier this week, Intel VP Kirk Skaugen released a PowerPoint slide detailing the rich history of Intel’s commitment to HPC innovation, its progression from the Petascale age to the Exascale age of supercomputing, and some hard specifications for its highly parallelized MIC architecture aimed at enterprise markets.

Back in 2007, we wrote that Larrabee was initially designed as a discrete graphics engine and was also capable of computing highly parallel applications while preserving x86 programmability. In May 2010, Bill Kircos, Intel’s Director of Product and Technology Media Relations, announced that the Larrabee project would never materialize as a discreet GPU part and would instead be transitioned into a new architecture leveraging both Larrabee and Intel’s many core research projects.

During ISC 2010, that architecture soon came to be known to the HPC crowd as Intel MIC (Many Integrated Core). In its official announcement, Intel outlined plans to ship a MIC development kit platform to select customers known as Knights Ferry. According to Slide 34 of Skaugen’s keynote presentation, Knights Ferry is an x86-based design with 32 cores on a single chip, each with four threads, a 32KB L1 instruction cache, a 32KB L1 data cache, and a 256KB L2 cache. In total, the chip has 8MB of shared L2 cache, which some analysts note to be an interesting design point as many high-parallel applications do not require such a large on-chip cache.

intel_knights_ferry_specs

Each processor has a very wide 512-bit vector unit allowing 16 single-precision floating point operations to be computed in a single instruction, with double-precision floating point operations yielding half throughput value.

Although the Knights Ferry development kit looks very similar to the outline of GPU, we are reminded to mention that it isn't a GPU because it has x86 cores. Besides, Intel would never do such a thing. Nevertheless, the card comes with a dual-slot heatsink, features up to 2GB of GDDR5 memory, and connects to a standard PCI-Express 2.0 motherboard slot. Intel advertises MIC as an “Intel Co-Processor Architecture,” so by nature it can become drop-in compatible with an Intel Xeon chip without the need to reprogram application code in another language.

intel_mic_architecture_programming

HPCwire.com has published a detailed architecture comparison between Intel’s Knights Ferry based on MIC architecture and Nvidia’s Tesla products based on Fermi architecture. As noted by Michael Wolfe, Slide 33 from Skaugen’s keynote presentation depicts the Knights Ferry architecture layout with remarkable similarity to the 2008 SIGGRAPH article describing Larrabee.

intel_larrabee_architecture_diagram

Figure 1: Schematic of the Larrabee many-core architecture: The number of CPU cores and the number and type of co-processors and I/O blocks are implementation-dependent, as are the positions of the CPU and non-CPU blocks on the chip.

intel_mic_architecture_schematic

Figure 2: Schematic of the Intel MIC (Many Integrated Core) architecture

Given the fact that Knights Ferry is not a commercially available product, it remains unclear whether or not it has similar design aspects to Knights Corner, the first MIC product Intel plans to launch. According to official plans, it will be manufactured on the 22nm half-pitch process node, will contain over 50 cores, and will be released sometime in 2011. All in all, we expect the next 16 months in the HPC sector to hold many interesting application performance competitions among Intel, AMD and Nvidia. While Intel boasts its x86 instruction set as a provider of maximum compatibility with existing applications without need for dual language programming on processor and co-processor, AMD and Nvidia focus their efforts on maximizing floating point throughput using a heterogeneous combination of CPUs with their Evergreen and Fermi GPU architectures.

intel_knights_family

Last modified on Wednesday, 11 August 2010 12:59
blog comments powered by Disqus

Comments  

 
+21 #1 DaRAGE 2010-08-11 12:58
Ok Look... I'm gunna say it....

But can it run Crysis?
:sigh:
 
 
+8 #2 Alexko 2010-08-11 13:22
No, because there are no drivers, but this is Larrabee, so with the proper drivers, it could.
 
 
+6 #3 FnuGk 2010-08-11 13:32
so this is like adding an extra cpu with lots of cores?

in other words it will make my rosetta@home score sky rocket?
 
 
-1 #4 blandead 2010-08-11 19:38
Quoting FnuGk:
so this is like adding an extra cpu with lots of cores?

in other words it will make my rosetta@home score sky rocket?


I THINK its just extra CPUs with 4x threads per CPU using GDDR5 mem. I don't see how this is better than a GPU where you can actually have 128 "cores, processors, shaders, whatever u wanna call em" not just 32 and multithreaded.

I dunno it's interesting, and could have advantages, but I still don't think it can replace GPU advantages if they incorporated it using 2 coding languages.
 
 
-1 #5 D31337Antics 2010-08-11 22:19
Intel is crap at making powerful GPUs so what they are trying to do is keep programmers lazy and keep the market buying their over priced CPUs.
 
 
-4 #6 trandoanhung1991 2010-08-12 01:04
Intel: IT'S ALIVE!!! MUAHAHAHAHAHAHA HA
 
 
+6 #7 Alereon 2010-08-12 03:02
Oh boy, 32 Atom-like processors with 4-way Hyperthreading...what are they targeting this for, tasks that are parallel enough to want more cores than a server, but not parallel enough to justify just recoding in CUDA/OpenCL to run on a Tesla/FireStream card? I'm not convinced that the HPC space is as married to x86 as Intel thinks it is, or that these simple x86 cores can provide competitive performance.
 
 
+1 #8 LuxZg 2010-08-12 08:52
Quoting blandead:
I THINK its just extra CPUs with 4x threads per CPU using GDDR5 mem. I don't see how this is better than a GPU where you can actually have 128 "cores, processors, shaders, whatever u wanna call em" not just 32 and multithreaded.


You don't get the two basic differences:
-this is x86 compatible and
-this is x86 compatible :)

Why did I say this twice? Well, first of all, you can't run ANYTHING on GPU, even if you worked your ass of to do it. Secondly, if you do want to run a part of tasks on GPU, you have to invest time & money in completely new application development. If only it was
 
 
+3 #9 LuxZg 2010-08-12 08:56
@Fudzilla staff - This comment script sucks bigtime.

To continue where I was cut off even though I had 100 characters left..

If only it was
 
 
+2 #10 LuxZg 2010-08-12 08:58
Quoting LuxZg:
@Fudzilla staff - This comment script sucks bigtime.

To continue where I was cut off even though I had 100 characters left..

If only it was


Fudzilla ... your comment section sucks. You should be checking for special characters like different parthensis types.. I can't even write a stupid "smaller than 300$" with simple sign, without getting cut off..

So I give up writing comment third time.. wasted enough time..
 

To be able to post comments please log-in with Disqus

 

Facebook activity

Latest Commented Articles

Recent Comments