Last Updated On: August 10, 2015
No discussion of the new DDR4 systems would be complete without taking a ride of Haswell-E lane. This article will explain what Haswell-E is, what features it inherited from Ivy Bridge, and what new features it boasts. This article will be fairly technical, so if you are not up to speed on the specifics I recommended checking out other articles on this site to get a good overview of what Haswell-E is and how it fits into the larger picture. I will try to keep this at a fairly high and abstract level, but be willing to bear with me on the occasional trudge through the technical goop.
Haswell-E is a processor architecture, much like the good old x286 and x386 processors from way back when. The architecture of a processor must match the architecture of the motherboard that it is being installed into for it the work. If they don’t match, you end up with a square peg in a round hole type of situation. This architecture is also the reason why you have to get a new processor if you upgrade your motherboard to take advantage of the new DDR4 memory. Most of the processors currently in use in computers do not support the Haswell-E architecture.
Different CPU architectures provide different features for the computer, as you will see below when we compare the Haswell architecture to the Ivy Bridge architecture. After that we will discuss the new features introduced by the Haswell- E architecture and how those features are designed to maximize the speed at which data moves through your computer. Those features on the Haswell-E architecture are what helps the computer take advantage of the performance enhancing aspects of the DDR4 motherboard and DDR4 memory platforms.
The release of the Haswell-E platform is good news for gamers and those with heavy workloads. The platform has support for eight core processors and up to four graphics cards, in addition to the DDR4 memory support. Unfortunately the Haswell-E platform only supports the Core i7 CPU, which means that you will have to upgrade both your motherboard and CPU before being able to use DDR4 memory.
The primary benefit that computer users can expect from the Haswell-E platform is to performance with lower power consumption. Now, the power consumption reduction is not going to make this new platform pay for itself. However, for users with heavy loads such as gamers, video editors, and developers, this platform packs plenty of punch for rendering, compiling, and playing. Finally, the Haswell-E platform processors can be overclocked up to 4.5 GHz which moves data a blinding speed.
If there is one consistent theme between the dd4 motherboards, memory, and processors it is overclocking. The motherboard BIOS makes it easy to overclock, and the Haswell-E platform provides a great overclocking foundation.
Haswell vs Ivy Bridge
The Haswell architecture kept several features from the Ivy Bridge architecture. These features include a 22 nm manufacturing process, 3D tri-gate transistors, micro-operation cache, 14- to 19-stage instruction pipeline, native support for DDR3 memory, and a total of 16 PCI Express lanes. Most computer users will have no idea what these features mean or what they do, and to be honest, these features are not the ones that really matter when looking at the Haswell-E system.
The key feature that carried over from Ivy Bridge to Haswell is the 16 PCI Express lanes. This allows for quite a bit of data to move from expansion cards such as video cards, surround sound audio cards, and network interface cards. In short, this allows for the computer to communicate with the extra features added by these cards faster and make these cards perform at a higher level.
New Haswell-E Features
One of the most obvious features that Haswell-E has over Ivy Bridge is the new sockets and chipsets that the architecture supports. For the desktop market, Haswell-E supports the LGA 2011-v3 along with the X99 chipset for the new DDR4 platform. This platform supports up to 8 cores and makes it easy to push the limits via overclocking.
Of course, these features are all overshadowed by the fact that Haswell-E is the first architecture to support DDR4 memory on the desktop platform. DDR4 memory brings higher speeds, faster clock speeds, and error correction features normally restricted to the server platform right to your desktop.
Another feature is the 20 MB shared cache and up to 40 PCIE lanes. These two features combine to move data at lightning fast speeds from the processor to memory to expansion cards. In practical terms, this means that video cards can receive instructions from the main processor and render graphics in memory faster, achieving higher frame rates at high resolution. For those not in the know, higher frame rates make animation appear much smoother and higher resolutions allow for sharper and more detailed images. Both of these outcomes are very desirable not only for gamers, but for multimedia artists, 3D artists, animators, and video creators.
Sounds great on paper, but what about reality?
All this performance enhancing talk about the Haswell-E platform makes it sound like Human Growth Hormone for your PC. But does the paper increases translate to real world performance? Bench mark data on systems built on both the Haswell-E architecture and the older DDR3 based architectures suggest that performance on 3D gaming, which is regarded as the most resource intensive application, is relatively similar.
This could be a side effect of the fact that this architecture is not yet in widespread usage. Therefore, game designers and application developers would be ill served in writing games and applications than can take full advantage of the performance of this platform. The goal of both game and application developers is to get as many users as possible for their product, and that is not achieved by writing to the cutting edge of hardware. The goal is best achieved by writing software that everyone can run, and at this point in the DDR4 life cycle, this means targeting the DDR3 platform until DDR4 has a higher adoption rate.
Should I upgrade?
The answer to that question depends on your expectations. If you are expecting life altering performance gains, you might be better off to wait until applications designed for the new platform are released. If you are expecting to replace your current system with a fairly future proof platform, then upgrading now isn’t the worse idea. Of course, you will have to pay a 60% financial premium to build a DDR4 based system over a comparable DDR3 system, but that if offset by being somewhat future proof against new application releases.
Wrapping it all up
Overall, the Haswell-E architecture is packed full of features designed to move data within your computer at much faster rates. This naturally leads to higher performance as data can reach all parts of your computer faster. From the moment you double-click a shortcut to the final rendering of the application your monitor, the Haswell-E architecture seeks to minimize the number of milliseconds that occur between these two events.
At the detail level the performance enhancing aspects of this platform look very promising. Faster data movement through the motherboard to the processor and back to the memory is always a good thing. However, practical performance is still relatively similar between the Haswell-E architecture and the older DDR3 systems because the adoption rate of the Haswell-E system is so low yet.
As time goes on and the adoption rate of the Haswell-E system climbs, application developers and game studios will start to target that platform. When that happens, the full power of the Haswell-E platform will be unlocked with richer graphics, smoother animations, and faster operations. Until that point in time, however, performance increases over the DDR3 based systems will be marginal at best.