Intel spent a lot of time talking about Nehalem a year ago, but not much time on Westmere. It's the tick to Nehalem's tock, or in other words, it's 32nm Nehalem.

Unlike previous die shrinks, we don't get larger caches with Westmere - Nehalem was already too big to begin with. Westmere keeps the same architecture, same cache sizes (or ratios) as Nehalem. It's all built using smaller 32nm transistors and on a smaller die. For the same core count, expect Westmere to be roughly half the size.

But the same core counts aren't what you're going to get. I included the table below in yesterday's Core i7 920XM preview:

The Westmere products are Gulftown, Clarkdale and Arrandale. That's six, two and two cores. Lynnfield is the last quad-core on the roadmap for the foreseeable future.

We'll talk about Gulftown later, but the focus today is Clarkdale with a little Arrandale.

Meet the 'dales

Arrandale and Clarkdale are the first two Westmere family members you'll meet. Both are technically due out later this year, although we won't see large volumes (by Intel standards) until Q1 2010. Both Arrandale and Clarkdale are dual-core Westmere parts with on-package graphics. The only difference is that Arrandale is mobile while Clarkdale is desktop.

Arrandale running - Hyper Threading helps improve performance even in normal workloads

The desktop socket is LGA-1156, the same socket as Lynnfield. The mobile socket is mPGA-989, the same socket as Clarksfield.

P55 vs. H57 Chipsets

The Lynnfield chipset is P55, it's the successor to P45 and the entire lineage of mainstream performance chipsets. The one thing P55 does not have is an integrated display interface; it has no way of connecting a CPU with on-package graphics, to a HDMI port (or VGA, or DVI).

The H57 chipset does:

Intel's display interface is a narrow bus that carries the video output from Clarkdale's on-package graphics and routes it to the video output on a motherboard.

DVI/HDMI out on a H57 mini-ITX motherboard

Why not give all chipsets video out support? Intel is big on differentiation. The P55 chipset is a bit more overclocking friendly, a set of optimizations that you won't see in H57. Both types of motherboards will take Clarkdale processors.

The First H57 mini-ITX Motherboard

Both micro-ATX and mini-ITX are being taken a lot more seriously now. With two-chip solutions commonplace on the desktop now, there's no reason we can't be building smaller motherboards:

Mini ITX H57 (left) vs. Mini ITX G45 (right)

You can expect this board to be available in Q1 2010.

Two Cores, Four Threads

Clarkdale is the spiritual successor to Conroe - one of our favorite dual-core processors of all time. It's a dual core chip but with all of the magic of Nehalem. Meaning you get a 64KB L1 per core, 256KB L2 per core but only a 4MB L3 cache. Remember that the L3 cache is shared among all cores and Intel likes to keep the ratio at 2MB per core.

Each core is capable of executing two threads (Hyper Threading). There's of course going to be turbo mode, although the upside shouldn't be too huge on desktop Clarkdale processors.

A Clarkdale mini-ITX system

Clarkdale, like the rest of the Nehalem/Westmere family makes extensive use of clock gating. You also get a ~1M transistor PCU that is in charge of keeping power consumption at a minimum. The result is a chip that truly sips power:

69.8W under Cinebench load

Intel's Clarkdale mini-ITX system used less than 30W at idle and only 70W under load running Cinebench's multi-threaded test. The idle performance is particularly impressive - that's not too far off from an Ion system honestly, but with much better performance.

27.6W at idle

AES-NI

Westmere does add some new instructions to x86, although the big expansion comes with AVX and Sandy Bridge next year. Westmere gets six new encryption/decryption instructions. The group of instructions accelerate AES (Advanced Encryption Standard) and are thus referred to as AES-NI.

All the new instructions need is software support and you'll see acceleration on Westmere. There were Winzip demos aplenty, which will have full acceleration support when the first Westmere chips hit.

When accelerated, Intel believes there will be no performance hit for encryption/decryption and it had the demos to back it up.

Here we're looking at disk throughput of an Intel SSD with full disk encryption enabled. Without AES-NI the performance takes a real hit:

40MB/s reads, 20 - 30MB/s writes

But with it, you can hardly tell you're even running AES:

~150MB/s reads, 50MB/s+ writes

Intel expects that full disk encryption and improved security will be commonplace once Westmere is in the mainstream.

On-package GPU and Graphics Turbo

Arrandale and Clarkdale are two-die packages. There's the 32nm CPU die and next to it is a 45nm DirectX 10 GPU die (no DX11 support until possibly Larrabee).

This isn't Larrabee (yet), it's a direct descendent of the graphics in G45. While G45 was built on a 65nm process, the 'dale graphics is built on a 45nm process.

The smaller transistors enable much higher performance. While G45 had 10 shader cores, the 'dale GPU increases that to 12. A number of performance limiting issues have now been resolved, so we should see much more competitive performance from Intel's graphics.

The memory controller has been moved off of the CPU die and is on the GPU die instead. It's still on-package so you get decently low latencies, but it shouldn't technically be as low as on Lynnfield. This is a temporary problem that fixes itself once the CPU/GPU are on the same die with Sandy Bridge.

Sandy Bridge brings on-die graphics

I've already explained turbo mode quite a bit so I won't rehash it here. The technology basically allows you to run your CPU at the fastest possible frequency regardless of how many cores are active. Westmere has this.

Arrandale will support graphics turbo modes, while Clarkdale won't. Clarkdale graphics is already running as fast as possible regardless of TDP.

If the GPU demand is higher than the CPU demand, the CPU will allocate more of its TDP to the GPU and vice versa.

Quad Core Performance From Two Cores?

Intel released some preliminary performance data on Clarkdale. This data wasn't run by us and is straight from Intel. It's normalized to a Core 2 Duo E8500 and also includes a Core 2 Quad Q9400. The systems used G45 graphics for the LGA-775 parts and Clarkdale on-package graphics for the LGA-1156 setup.

We don't usually look at SPEC CPU performance, but here are the results regardless:

You can see that thanks to a competitive clock speed, aggressive turbo modes and Hyper Threading the 3.33GHz Clarkdale outperforms both the Q9400 and the E8500.

3DMark Vantage is of course faster on the newer platform thanks to the on-package graphics:

The overall performance improvement is roughly 50% in 3DMark Vantage.

Memory bandwidth is clearly better thanks to Clarkdale's on-die dual-channel DDR3 memory controller.

PCMark Vantage also shows a huge performance advantage on Clarkdale, however Intel mentioned that a big portion of that are the AES-NI instructions on Clarkdale accelerating one of the tests.

We'll be testing Clarkdale for ourselves later this year. It seems that this potent dual core with Hyper Threading could be a good alternative to making the quad-core jump.

Bitstreaming TrueHD/DTS-HD MA: Yep, Here too

The Radeon HD 5870 was the first graphics card to properly support bitstreaming of high definition Blu-ray audio codecs. Clarkdale/Arrandale is the second.

These CPUs come with an on-package GPU and that GPU supports the appropriate protected audio path to enable bitstreaming of Dolby TrueHD and DTS-HD MA. Of course 8-channel LPCM output is also still an option.

If you remember the G45 launch, Intel had serious issues enabling 8-channel LPCM output, HDCP and H.264 decode acceleration in general. I grilled Intel on what was going to make this round different and they are much more confident in their abilities.

They've increased the number of receivers they test with (originally it was at a whopping two, now they're up to…7). They've also expanded their test scenarios as well. The combination of the two, Intel believes, will result in a fully functional set of HTPC features at launch.

The first time I went by Intel's Clarkdale demo, Intel couldn't get bitstreaming working. A day later I got an email telling me to drop by again - they fixed it.

I got to see TrueHD bitstreaming from a Clarkdale system to a Sony receiver. I also confirmed that full two stream decode acceleration was working:

Intel had it working with Arcsoft's player, but is working with all of the major software vendors to hopefully enable full support on everything. Intel does seem to be taking this much more seriously than with G45.

The Clarkdale launch is still a couple of months away so there is definitely time for Intel to work out the kinks.

This is a serious feature. The fact is that in a couple of years every single PC shipped will have the ability to bitstream these audio codecs without any additional hardware. We're finally getting there folks.

Gulftown: 6-cores, 32nm, Backwards Compatible with X58

The final bit on Westmere is Gulftown. This is a LGA-1366 processor with six cores, 12MB of L3 cache and fully backwards compatible with X58 motherboards (provided your motherboard manufacturer enables it).

Westmere isn't going to be a huge upgrade for the quad-core space, but if you've got an older dual-core system this is going to be the generation to upgrade to.