Now that Core i7 is in the market, and we have tested few machines that run on Whooping i7, we are all set to write another review on Core i7. As we mentioned in our initial review of the Intel Core i7 Nehalem platform, the 920 and 940 run with different speeds than the 965 – beyond just the clock speed. The actual memory clocks run at different speeds, such as the QPI (QuickPath Interface, which replaces the FSB), L3 cache, and even the speed of the DDR3 memory being used.

The 965 normally has a QPI frequency of 6.4 GT/s, with the memory interface running at 1600 MHz. The 940 and 920 are meant to run with a 4.8 GT/s QPI and 1066 MHz memory interface.

We were told by Intel that these were all locked, and there would be no way to change this, without raising the 133 MHz base clock, which as we know, can significantly impact stability.

Naturally, all our Core i7 CPU’s are engineering samples from Intel, which are completely locked. But people are now discovering that their retail Core i7s have completely unlocked QPI and memory interface speeds.

As you probably know by now the Intel Core series of processors is not just a rehash of its previous Core 2 Duo technology. Instead it uses a completely unique design that embraces much of the new Nehalem technology fully described in “First the Tick, Now the Tock: Next Generation IntelMicroarchitecture (Nehalem)”. Intel has gone to great lengths to avoid having their newest product entitled “just another revision” by their consumer market. They have shown their confidence in the Intel Core series by making this product launch the biggest and arguably the best in their history, even in a time where the economy is less than prosperous.

So as long as you have a decent motherboard that can manually set these speeds, you can enjoy the same snappy QPI and full speed DDR3. Yup, the Core i7 already kicked ass, but now it’s even better than we thought. Checkout all the benchmarks below (courtesy Tom’s hardware, PCWorld, gizmodo, HardcoreWare):

Nehalem, 3DMark06 CPU Score : Nehalem, 3DMark06 CPU Score:

Nehalem, Everest Cache and Memory Benchmark @ Dual Channel : Nehalem, Everest Cache and Memory Benchmark @ Dual Channel:

Nehalem, Everest Cache and Memory Benchmark @ Triple Channel : Nehalem, Everest Cache and Memory Benchmark @ Triple Channel:

Nehalem 3DStudio Max 8 Rendering : Nehalem 3DStudio Max 8 Rendering:

Nehalem, Canopus Pro Coder 3 : Nehalem, Canopus Pro Coder 3:

Nehalem, Cinebench R10 Multicore bench : Nehalem, Cinebench R10 Multicore bench:

Specifications:

With the Core i7 series Intel has produced their first true, native quad-core processor which incorporates multiple cores on a single processor die. This is unlike the existing Core 2 Quads which feature two dual core processor dies mounted on a single piece of silicon. The Core i7 processors also eliminate the need for the Northbridge chipset to manage system memory and link it to the processor with integrated, on die memory controller which links the processor directly to the system memory. You might be asking yourself: Isn’t this design is eerily similar to the AMD Phenom? From this limited point of reference you would be right.

Intel Core i7 Features

Another very interesting feature is Intel’s QuickPath Interface. QuickPath Architecture is a platform architecture that provides high-speed connections between microprocessors and external memory, and between microprocessors and the I/O hub. One of its biggest changes is the implementation of scalable shared memory. Instead of using a single shared pool of memory connected to all the processors in a server or high-end workstation through FSBs and memory controller hubs, each processor has its own dedicated memory that it accesses directly through an Integrated Memory Controller on the processor die. In cases where a processor needs to access the dedicated memory of another processor, it can do so through a high-speed Intel QuickPath Interconnect that links all the processors. An advantage of Intel QuickPath Interconnect is that it is point-to-point. There is no single bus that all the processors must use and contend with each other to reach memory and I/O. This improves scalability and eliminates the competition between processors for bus bandwidth.