Intel Westmere 32nm Launch & Clarkdale Core i5-661 CPU Review

Westmere Microarchitecture - Clarkdale Edition

In 2007, Intel unveiled the Tick-Tock Model as a demonstration of the company's dedication towards continued rapid technological innovation. The "tick" is a shrinking of the previous architecture manufacturing process (65nm --> 45nm --> 32nm) and the "tock" is a new architecture. First launched in November 2008 with the Bloomfield Core i7-900 series, the Nehalem microarchitecture saw its second variant launched in September 2009; the Lynnfield Core i5-700 series & i7-800 series. Lynnfield was nothing radically new, it merely used Nehalem's modular design to integrate the PCI-Express controller into the CPU die.

Today, with this 2010 launch, we finally have the "tick" to 2008's "tock". Westmere shrinks the Nehalem architecture down to 32nm, which is basically a 41% smaller manufacturing process than 45nm. This is important because manufacturing cores on a smaller process allows for cooler running chips, better power efficiency, higher frequency scaling, more cores on a CPU package, and most importantly it allows for Nehalem-based parts to be introduced at lower priced points, which is what Clarkdale is all about.

Enthusiasts needn't worry though since Clarkdale is just a budget-friendly introduction to Westmere, we will be seeing proper 32nm quad-core and six-core Gulftown chips sooner than you might think. Further down the road, the brand new Sandy Bridge microarchitecture will also be manufactured on this new 32nm high-k + metal gate transistor technology.



The integrated graphics processor (IGP) that is found on Clarkdale chips is not part of the Westmere microarchitecture, so we will take a look at it in another section. What we are interested in is the smaller of the two dies, which is the 32nm dual-core CPU itself. Clarkdale's core design is unique, hence the Clarkdale Edition sub-title since no other Westmere-based processors will be similar to this variant.



Despite being based on the same microarchitecture, the Clarkdale and Lynnfield dies are quite different. One would reasonably expect Clarkdale to simply be half of a Lynnfield, but the situation is a little bit more complicated than that. While the basic blocks of the die are effectively identical, Clarkdale does away with the integrated memory controller and the PCI-Express controller, which have been relocated to the Ironlake Graphics Memory Controller Hub (GMCH), ie: the 45nm die on the CPU package which also contains the IGP. This not only makes the Clarkdale CPU die easier and cheaper to manufacture, but it also gives the IGP direct access to the system memory and PCI-E controller for low latency communication. What effect this will have on the CPU's access to the system memory will be explored in the memory bandwidth section.

In quantifiable terms, Clarkdale's CPU die size is a tiny 81mm². By comparison, Lynnfield comes in at 296mm², Bloomfield measures 263mm², and the Core 2 Quad "Yorkfield" die is 214mm². That itty-bitty little CPU die is packed with 383 million transistors, which is fairly amazing when you compare it to Lynnfield's 774 million transistors and Bloomfield's 731 million transistors. It is not entirely a fair comparison though since the Clarkdale CPU die does lack the aforementioned integrated memory controller, PCI-E controller, and has half the L3 cache. Nevertheless, the advantages of the new 32nm manufacturing process are evident.

Now many of you are probably looking at the die pictures and saying "It is pretty but what am I looking at exactly?". A valid question, so let's take a look at the Clarkdale CPU core layout:


We are a few PhD's away from being able to critique the core layout, but just admire how simple this die is compared to Lynnfield's.


Part of the reason that Intel is able to add and remove parts so easily is because the Nehalem architecture is dynamically scalable, and it was designed with modularity in mind. What this means is that Intel can custom create processors based on the needs of the market without having to go design a brand new chip from scratch. They can add or remove cores, L3 cache, memory channels, memory controllers, power management features, and even integrated graphics. Therefore, Intel have the ability to add new blocks to the core without having to go to the drawing board and redesigning the whole layout. Basically, they are only limited by how much stuff they can actually fit on one CPU package. Think of it as a multi-million dollar Lego set.


If you are unfamiliar with the features and technology present in Nehalem-based processors, or simply want a more in-depth explanation as to what Intel have done to Clarkdale's CPU die, the following section should interest you.