[H]ard|OCP

Navigation

• [H] Forums
• Archives
• Search
• [H] Folding

Intel 3.06GHz CPU with Hyper-Threading

Intel bangs the 3GHz barrier before anyone else and throws in HT for free. We know 3GHz is going to be flying, but the world really wants to know all about HT and what it has to offer.

Today we are looking at the 3.06GHz Hyper-Threading enabled CPU from Intel. The CPU itself is a milestone for Intel simply in the fact that it has broken the 3GHz barrier like they promised it would over a year ago. There is, however, something more that makes this processor launch different from all the other ones. What makes this processor really special can be summarized in one word: Hyper-Threading. What is Hyper-Threading and what does it mean to you? Trust us, we wanted to know too.

I want to preface this review by saying that what will make this review somewhat different from almost all of the others we've done is in our conclusion, and how we reach it. A standard review is, and should be, free from basing conclusions solely on opinion or simply “what we think”. Conclusions are generally based on hard facts like performance numbers derived from a set of industry standard tests and a product's stability over time. We always add a “dash” of our own opinion on how we feel the product will effect the industry and end users as well, but raw performance data usually dictates the outcome of a review, even over our own opinions of a product.

After extensively testing the new 3.0GHz Hyper-Threading enabled processor, I believe that what we feel about the processor and what it brings to the table will weigh equally to the results we saw throughout our benchmarking process. First, subjective VidCard reviews, now CPUs. What is the world coming to?

What is Hyper-Threading?

From Intel:

The amazing growth of the Internet and telecommunications is powered by ever-faster systems demanding increasingly higher levels of processor performance. To keep up with this demand we cannot rely entirely on traditional approaches to processor design. Microarchitecture techniques used to achieve past processor performance improvement – super-pipelining, branch prediction, super-scalar execution, out-of-order execution, caches – have made microprocessors increasingly more complex, have more transistors, and consume more power. In fact, transistor counts and power are increasing at rates greater than processor performance. Processor architects are therefore looking for ways to improve performance at a greater rate than transistor counts and power dissipation. Intel's Hyper-Threading Technology is one solution.

Hyper-Threading is a technology developed by Intel to make a single Pentium4 processor appear, for all intents and purposes, as two processors to the operating system and subsequent applications. When we think of programs, applications, and games, we rarely think of them in the form of threads, or strings of data, fed to the processor to produce what we actually see on our monitor.

In the example we see above, a single processor system must handle multiple tasks one of two ways. The CPU can either take each task in order, one after the other and process them, or it can process multiple tasks by alternating which between processes are handled. Both approaches result in the data being processed in separate manners and yielding the same results, because the data is fed to the processor a thread at a time. The final example shows Hyper-Threading technology at work, simultaneously processing dual tasks and allowing both tasks to be completed at once, instead of having to wait for one to finish first. The example above is a bit extreme in showing the time savings, but it was done that way to give you an understanding of the process.

These threads of information, when processed by a single conventional CPU, can only be processed one at a time in, excuse the term, “single file” fashion. The Hyper-Threading enabled Pentium4 has been modified in such a way that multiple threads can execute on one processor, much in the same way multiprocessor configurations are able to process data through switching data between processors. In the case of Hyper-Threading, multiple threads are able to execute simultaneously without switching, which makes much better use of the resources by carrying out all processes on a single CPU.

Hyper-Threading does not mean that there are two physical processor cores sharing the same CPU. In fact, there is very little in actual physical change to the processor. Studying the CPU die graph above, you can see where key components responsible for enabling Hyper-Threading are located on the actual processor die.

An explanation of what you're seeing is summed up here:

Each logical processor maintains a complete set of the architecture state. The architecture state consists of registers including the general-purpose registers, the control registers, the advanced programmable interrupt controller (APIC) registers, and some machine state registers. From a software perspective, once the architecture state is duplicated, the processor appears to be two processors. The number of transistors to store the architecture state is an extremely small fraction of the total. Logical processors share nearly all other resources on the physical processor, such as caches, execution units, branch predictors, control logic, and buses.