After three long years of going hungry with quad-cores, red meat is finally back on the menu for enthusiasts. And not just any gamey slab full of gristle with shared cores, either. With its new eight-core Haswell-E CPU, Intel may have served up the most mouth-watering, beautifully seared piece of red meat in a long time.
And it’s a good thing, too, because enthusiast’s stomachs have been growling. Devil’s Canyon? That puny quad-core was just an appetizer. And that dual-core highly overclockable Pentium K CPU? It’s the mint you grab on your way out of the steak house.
No, what enthusiasts have craved and wanted ever since Intel’s original clock-blocking job on the original Sandy Bridge-E was a true, overclockable enthusiast chip with eight cores. So if you’re ready for a belt loosening, belly full of enthusiast-level prime rib, pass the horse radish, get that damned salad off our table, and read on to see if Intel’s Haswell-E is everything we hoped it would be.
Meet the Haswell-E parts
Despite its name, the LGA2011-v3 socket is not same as the older LGA2011 socket. Fortunately, the cooling offsets are exactly the same, so almost all older coolers and accessories should work just fine.
Though they look the same, LGA2011’s socket has arms that are actually arranged differently than the new LGA2011-v3 that replaces it. And no, you can’t drop a newer Haswell-E into this socket and make it work.
The first consumer Intel eight-core arrives at last
Being a card-carrying member of the PC enthusiast class is not an easy path to follow. Sure, you get the most cores and priciest parts, but it also means you get to wait a hell of a long time in between CPU upgrades. And with Intel’s cadence the last few years, it also means you get the leftovers. It’s been that way ever since Intel went with its two-socket strategy with the original LGA1366/LGA1156. Those who picked the big-boy socket and stuck to their guns on Pure PC performance always got the shaft.
The original Ivy Bridge in LGA1156 socket, for example, hit the streets in April of 2012. As a reward for having the more efficient and faster CPU, Intel rewarded the small-socket crowd with its Haswell in June of 2013. It wasn’t until September of 2013 that big-boy socket users finally got Ivy Bridge-E for their LGA2011s. But with Haswell already out and tearing up the benchmarks, who the hell cared?
Well, that time has come with Haswell-E, Intel’s first replacement for the aging LGA2011 platform since 2011. This time though, Intel isn’t just shuffling new parts into its old stack. For the first since the original Pentium 4 Extreme Edition, paying the price premium actually nets you more: namely, the company’s first consumer eight-core CPU.
Meet the T-Rex of consumer CPUs: The Core i7-5960X
We were actually a little leery of Haswell when it first launched last year. It was, after all, a chip seemingly tuned for the increasingly mobile/laptoppy world we were told was our post PC-apocalyptic future. Despite this, we recognized the chip as the CPU to have for new system builders. Clock for clock, its 22nm process, tri-gate transistors put everything else to shame—even the six-core Core i7-3930K chip in many tasks. So it’s no surprise that when Intel took a quad-core Haswell, put it in the Xerox machine, and hit the copy x2 button , we’d be ecstatic. Eight cores are decidedly better than six cores or four cores when you need them.
The cores don’t come without a cost though, and we don’t mean the usual painful price Intel asks for its highest-end CPUs. It’s no secret that more cores means more heat, which means lower clock speeds. That’s one of the rationales Intel used with the original six-core Core i7-3960X. Although sold as a six-core, the original Sandy Bridge-E was built using an eight-core die on which Intel had permanently switched off two cores. Intel said it wanted to balance the needs of the many versus the needs of the few—that is, by turning off two of the cores, the part could hit higher clock speeds. Indeed, the Core i7-3960X had a base clock of 3.3GHz and Turbo Boost of 3.9GHz, and most could overclock it to 5GHz. The same chip packaged as a Xeon with all eight cores working—the Xeon E5-2687W—was locked down at 3.1GHz and mostly buzzed along at 3.4GHz.
With the new Core i7-5960X—the only eight-core of the bunch—the chip starts at a seemingly pedestrian 3GHz with a Turbo Boost of one core up to 3.5GHz. Those subsonic clock speeds won’t impress against the Core i7-4790K, which starts at 4GHz. You’ll find more on how well Haswell-E performs against Haswell in our performance section, but that’s the price to be paid, apparently, to get a chip with this many cores under the heat spreader. Regarding thermals, in fact, Intel has increased the TDP rating to 140 watts versus 130 watts of Ivy Bridge-E and Sandy Bridge-E.
If the low clocks annoy you, the good news is the part is fully unlocked, so the use of overclocking has been approved. For our test units, we had very early hardware and tight deadlines, so we didn’t get very far with our overclocking efforts. Talking with vendors, however, most seem very pleased with the clock speeds they were seeing. One vendor told us overclocks of all cores at 4.5GHz was already obtainable and newer microcode updates were expected to improve that. With even the vaunted Devil’s Canyon Core i7-4790K topping out at 4.7GHz to 4.8GHz, a 4.5GHz is actually a healthy overclock for an eight-core CPU.
When you dive down into the actual cores though, much is the same, of course. It’s based on a 22nm process. It has “3D” tri-gate transistors and integrated voltage regulation. Oh, and it’s also the first CPU to feature an integrated DDR4 memory controller.
Click the next page to read about DDR4
If you think Haswell-E has been a long wait, just think about DDR3, which made its debut as main memory in systems since 2007. Yes, 2007. The only component that has lasted seven years in most enthusiasts systems might be the PSU, but it’s even rare to find anyone kicking a 500-watt PSU from 2007 these days.
DDR4 has been in gestation seemingly as long, so why the delay? From what we can tell, resistance to yet another new memory standard during a time when people thought the desktop PC and the PC in general were dying has been the root delay. It didn’t help that no one wanted to stick their head out first, either. RAM makers didn’t want to begin producing it DDR4 in volume until AMD or Intel made chipsets for it, and AMD and Intel didn’t want to support it because of the costs it would add to PCs at a time when people were trying to lower costs. The stalemate finally ends with Haswell-E, which integrates a quad-channel memory controller into its die.
Initial launch speeds of DDR4 clock in at DDR4/2133. For those already running DDR3 at 3GHz or higher, a 2,133 data rate is a snooze, but you should realize that anything over 2133 is overclocked RAM. With DDR4, the JEDEC speeds (the body that sets RAM standards) already has target data rates of 3200 on the map. RAM vendors we’ve talked to are already shopping DIMMS near that speed.
The best part of DDR4 may be its density message, though. For years, consumer DDR3 has topped out at 8GB on a DIMM. With DDR4, we should see 16B DIMMs almost immediately, and stacking of chips is built into the standard, so it’s possible we’ll see 32GB DIMMs over its lifetime. On a quad-channel, eight-DIMM motherboard, you should expect to be able to build systems with 128GB of RAM using non-ECC DIMMs almost immediately. DDR4 also brings power savings and other improvements, but the main highlights enthusiasts should expect are higher densities and higher clocks. Oh, and higher prices. RAM prices haven’t been fun for anyone of late, but DDR4 will definitely be a premium part for some time. In fact, we couldn’t even get exact pricing from memory vendors as we were going to press, so we’re bracing for some really bad news.
Over the years, we’ve come to expect Intel to clock-block core counts, clock speeds, Hyper-Threading, and even cache for “market segmentation” purposes. What that means is Intel has to find ways to differentiate one CPU from another. Sometimes that’s by turning off Hyper-Threading (witness Core i5 and Core i7) and sometimes its locking down clock speeds. With Haswell-E though, Intel has gone to new heights with its clock-blocking by actually turning off PCIe lanes on some Haswell-E parts to make them less desirable. At the top end, you have the 3GHz Core i7-5960X with eight cores. In the midrange you have the six-core 3.5GHz Core i7-5930K. And at the “low-end” you have the six-core 3.3GHz Core i7-5820K. The 5930K and the 5820K are virtually the same in specs except for one key difference: The PCIe lanes get blocked. Yes, while the Core i7-5960X and Core i7-5930K get 40 lanes of PCIe 3.0, the Core i7-5820K gets an odd 28 lanes of PCIe 3.0. That means those who had hoped to build “budget” Haswell-E boxes with multiple GPUs may have to think hard and fast about using the lowest-end Haswell-E chip. The good news is that for most people, it won’t matter. Plenty of people run Haswell systems with SLI or CrossFire, and those CPUs are limited to 16 lanes. Boards with PLX switches even support four-way GPU setups.
Still, it’s a brain bender to think that when you populate an X99 board with the lowest-end Haswell-E, the PCIe configuration will change. The good news is at least they’ll work, just more slowly. Intel says it worked with board vendors to make sure all the slots will function with the budget Haswell-E part.
There have been clock-blocking rumors swirling around about the Haswell being a 12-core Xeon with four cores turned off. That’s not true and Intel says this die-shot proves it.
Ivy Bridge-E’s main advantage over Sandy Bridge-E was a native six-core die and greatly reduced power consumption. And, unfortunately, like its Ivy Bridge counterpart, overclocking yields on Ivy Bridge-E were greatly reduced over its predecessor, too, with few chips hitting more than 4.7GHz at best.
Sandy Bridge-E and Sandy Bridge will long be remembered for its friendliness to overclocking and having two of its working cores killed Red Wedding–style by Intel.
Click the next page to read about X99.
High-end enthusiasts finally get the chipset they want, sort of
Intel overcompensated in SATA on X99 but oddly left SATA Express on the cutting-room floor.
You know what we won’t miss? The X79 chipset. No offense to X79 owners, while the Core i7-4960X can stick around for a few more months, X79 can take its under-spec’ed butt out of our establishment. Think we’re being too harsh? We don’t.
X79 has no native USB 3.0 support. And its SATA 6Gb/s ports? Only two. It almost reads like a feature set from the last decade to us. Fortunately, in a move we wholly endorse, Intel has gone hog wild in over-compensating for the weaknesses of X79.
X99 has eight USB 2.0 ports and six USB 3.0 ports baked into the peripheral controller hub in it. For SATA 6Gb/s, Intel adds 10 ports to X99. Yes, 10 ports of SATA 6Gb/s. That gazongo number of SATA ports, however, is balanced out by two glaring omission in X99: no official SATA Express or M.2 support that came with Z97. Intel didn’t say why it left off SATA Express or M.2 in the chipset, but it did say motherboard vendors were free to implement it using techniques they gleaned from doing it on Z97 motherboards. If we had to hazard a guess, we’d say Intel’s conservative nature led it to leave the feature off the chipset, as the company is a stickler for testing new interfaces before adding official support. At this point, SATA Express has been a no-show. After all, motherboards with SATA Express became available in May with Z97, yet we still have not seen any native SATA Express drives. We expect most motherboard vendors to simply add it through discrete controllers; even our early board sample had a SATA Express port.
One potential weakness of X99 is Intel’s use of the DMI 2.0. That offers roughly 2.5GB/s of transfer speed between the CPU and the south bridge or PCH, but with the board hanging 10 SATA devices, USB 3.0, Gigabit Ethernet, and 8 PCIe Gen 2.0 lanes off that link, there is the potential for massive congestion—but only in a worst-case scenario. You’d really have to a boat load of hardware lit up and sending and receiving data at once to cause the DMI 2.0 to bottleneck. Besides, Intel says, you can just hang the device off the plentiful PCIe Gen 3.0 from the CPU.
That does bring up our last point on X99: the PCIe lanes. As we mentioned earlier, there will be some confusion over the PCIe lane configuration on systems with Core i7-5820K parts. With only 28 lanes of PCIe lanes available from that one chip, there’s concern that whole slots on the motherboard will be turned off. That won’t happen, Intel says. Instead, if you go with the low-rent ride, you simply lose bandwidth. Take an X99 mobo and plug in the Core i7-5930K and you get two slots at x16 PCIe, and one x8 slot. Remove that CPU and install the Core i7-5820K, and the slots will now be configured as one x16, one x8 and one x4. It’s still more bandwidth than you can get from a normal LGA1150-based Core i7-4770K but it will be confusing nonetheless. We expect motherboard vendors to sort it out for their customers, though.
Haswell-E does bring one more interesting PCIe configuration though: the ability to run five graphics cards in the PCIe slots at x8 speeds. Intel didn’t comment on the reasons for the option but there only a few apparent reasons. The first is mining configurations where miners are already running six GPUs. Mining, however, doesn’t seem to need the bandwidth a x8 slot would provide. The other possibility is a five-way graphics card configuration being planned by Nvidia or AMD. At this point it’s just conjecture, but one thing we know is that X99 is a welcome upgrade. Good riddance X79.
How many cores do you really need?
It is indeed a glorious thing to see a task manager with this many threads, but not everyone needs them.
Like the great technology philosopher Sir Mix-A-Lot said, we like big cores and we cannot lie. We want as many cores as legally available. But we recognize that not everyone rolls as hard as we do with a posse of threads. With Intel’s first eight-core CPU, consumers can now pick from two cores all the way to eight on the Intel side of the aisle—and then there’s Hyper-Threading to confuse you even more. So, how many cores do you need? We’ll give you the quick-and-dirty lowdown.
Normally, we’d completely skip dual-cores without Hyper-Threading because the parts tend to be the very bottom end of the pool Celerons. Our asterisk is the new Intel Pentium G3258 Anniversary Edition, or “Pentium K,” which is a real hoot of a chip. It easily overclocks and is dead cheap. It’s not the fastest in content creation by a long shot, but if we were building an ultra-budget gaming rig and needed to steal from the CPU budget for a faster GPU, we’d recommend this one. Otherwise, we see dual-cores as purely ultra-budget parts today.
Two cores with Hyper-Threading
For your parents who need a reliable, solid PC without overclocking (you really don’t want to explain how to back down the core voltage in the BIOS to grandma, do you?), the dual-core Core i3 parts fulfill the needs of most people who only do content creation on occasion. Hyper-Threading adds value in multi-threaded and multi-tasking tasks. You can almost think of these chips with Hyper-Threading as three-core CPUs.
For anyone who does content creation such as video editing, encoding, or even photo editing with newer applications, a quad-core is usually our recommended part. Newer game consoles are also expected to push min specs for newer games to quad-cores or more as well, so for most people who carry an Enthusiast badge, a quad-core part is the place to start.
Four cores with Hyper-Threading
Hyper-Threading got a bad name early on from the Pentium 4 and existing software that actually saw it reduce performance when turned on. Those days are long behind us though, and Hyper-Threading offers a nice performance boost with its virtual cores. How much? A 3.5GHz Core i7 quad-core with Hyper-Threading generally offers the same performance on multi-threaded tasks as a Core i5 running at 4.5GHz. The Hyper-Threading helps with content creation and we’d say, if content creation is 30 percent or less of your time, this is the place to be and really the best fit for 90 percent of enthusiasts.
Six cores with Hyper-Threading
Once you pass the quad-core mark, you are moving pixels professionally in video editing, 3D modeling, or other tasks that necessitate the costs of a six-core chip or more. We still think that for 90 percent of folks, a four-core CPU is plenty, but if losing time rendering a video costs you money (or you’re just ADD), pay for a six-core or more CPU. How do you decide if you need six or eight cores? Read on.
Eight cores with Hyper-Threading
We recognize that not everyone needs an eight-core processor. In fact, one way to save cash is to buy the midrange six-core chip instead, but if time is money, an eight-core chip will pay for itself. For example, the eight-core Haswell-E is about 45 percent faster than the four-core Core i7-4790K chip. If your render job is three hours, that’s more time working on other paying projects. The gap gets smaller between the six-core and the eight-core of course, so it’s very much about how much your time is worth or how short your attention span is. But just to give you an idea, the 3.3GHz Core i7-5960X is about 20 percent faster than the Core i7-4960X running at 4GHz.
Click the next page to see how Haswell-E stacks up against Intel's other top CPUs.
The LGA2011-based Core i7-4960X (left) and the LGA2011-v3-based Core i7-5960X (middle) dwarf the Core i7-4790K chip (right). Note the change in the heat spreader between the older 4960X and 5960X, which now has larger “wings” that make it easier to remove the CPU by hand. The breather hole, which allows for curing of the thermal interface material (solder in this case), has also been moved. Finally, while the chips are the same size, they are keyed differently to prevent you from installing a newer Haswell-E into an older Ivy Bridge-E board.
Performance junkies, rejoice! Haswell-E hits it out of the ballpark
We used a Gigabyte X99 motherboard (without the final heatsinks for the voltage-regulation modules) for our testing.
For our testing, we set up three identical systems with the fastest available CPUs for each platform. Each system used an Nvidia GeForce GTX 780 with the same 340.52 drivers, Corsair 240GB Neutron GTX SSDs, and 64-bit Windows 8.1 Enterprise. Since we’ve had issues with clock speeds varying on cards that physically look the same, we also verified the clock speeds of each GPU manually and also recorded the multiplier, bclock, and speeds the parts run at under single-threaded and multi-threaded loads. So you know, the 3GHz Core i7-5960X’s would run at 3.5GHz on single-threaded tasks but usually sat at 3.33GHz on multi-threaded tasks. The 3.6GHz Core i7-4960X ran everything at 4GHz, including multi-threading tasks. The 4GHz Core i7-4790K part sat at 4.4GHz on both single- and multi-threaded loads.
For Z97, we used a Gigabyte Z97M-D3H mobo with a Core i7-4790K “Devil’s Canyon” chip aboard. An Asus Sabertooth X79 did the duty for our Core i7-4960X “Ivy Bridge-E” chip. Finally, for our Core i7-5960X chip, we obtained an early Gigabyte X99-Gaming 5 motherboard. The board was pretty early but we feel comfortable with our performance numbers as Intel has claimed the Core i7-5960X was “45 percent” faster than a quad-core chip, and that’s what we saw in some of our tests.
One thing to note: The RAM capacities were different but in the grand scheme of things and the tests we run, it has no impact. The Sabertooth X79 had 16GB of DDR3/2133 in quad-channel mode, the Z97M-D3H had 16GB of DDR3/2133 in dual-channel mode. Finally, the X99-Gaming 5 board had 32GB of Corsair DDR4/2133. All three CPUs will overclock, but we tested at stock speeds to get a good baseline feel.
For our benchmarks, we selected from a pile of real-world games, synthetic tests, as well as real-world applications across a wide gamut of disciplines. Our gaming tests were also run at very low resolutions and low-quality settings to take the graphics card out of the equation. We also acknowledge that people want to know what they can expect from the different CPUs at realistic settings and resolutions, so we also ran all of the games at their highest settings at 1920x1080 resolution, which is still the norm in PC gaming.
We could get into a multi-sentence analysis of how it did and slowly break out with our verdict but in a society where people get impatient at the microwave, we’ll give you the goods up front: Holy Frakking Smokes, this chip is fast! The Core i7-5960X is simply everything high-end enthusiasts have been dreaming about.
Just to give you an idea, we’ve been recording scores from $7,000 and $13,000 PCs in our custom Premiere Pro CS6 benchmark for a couple of years now. The fastest we’ve ever seen is the Digital Storm Aventum II that we reviewed in our January 2014 issue. The 3.3GHz Core i7-5960X was faster than the Aventum II’s Core i7-4960X running at 4.7GHz. Again, at stock speeds, the Haswell-E was faster than the fastest Ivy Bridge-E machine we’ve ever seen.
It wasn’t just Premiere Pro CS6 we saw that spread in either. In most of our tests that stress multi-threading, we saw roughly a 45 percent to 50 percent improvement going from the Haswell to the Haswell-E part. The scaling gets tighter when you’re comparing the six-core Core i7-4960X but it’s still a nice, big number. We generally saw a 20 percent to 25 percent improvement in multi-threaded tasks.
That’s not even factoring in the clock differences between the parts. The Core i7-4790K buzzes along at 4.4GHz—1.1GHz faster than the Core i7-5960X in multi-threaded tasks—yet it still got stomped by 45 to 50 percent. The Core i7-4960X had a nearly 700MHz clock advantage as well over the eight-core chip.
The whole world isn’t multi-threaded, though. Once we get to workloads that don’t push all eight cores, the higher clock speeds of the other parts predictably take over. ProShow Producer 5.0, for example, has never pushed more than four threads and we saw the Core i7-5960X lose by 17 percent. The same happened in our custom Stitch.Efx 2.0 benchmark, too. In fact, in general, the Core i7-4790K will be faster thanks to its clock speed advantage. If you overclocked the Core i7-5960X to 4GHz or 4.4GHz on just four cores, the two should be on par in pure performance on light-duty workloads.
In gaming, we saw some results from our tests that are a little bewildering to us. At low-resolution and low-quality settings, where the graphics card was not the bottleneck, the Core i7-4790K had the same 10 percent to 20 percent advantage. When we ran the same tests at ultra and 1080p resolution, the Core i7-5960X actually had a slight advantage in some of the runs against the Core i7-4790K chip. We think that may be from the bandwidth advantage the 5960X has. Remember, we ran all of the RAM at 2,133, so it’s not DDR4 vs. DDR3. It’s really quad-channel vs. dual-channel.
We actually put a full breakdown of each of the benchmarks and detailed analysis on MaximumPC.com if you really want to nerd out on the performance.
What you should buy
Let’s say it again: The Core i7-5960X stands as the single fastest CPU we’ve seen to date. It’s simply a monster in performance in multi-threaded tasks and we think once you’ve overclocked it, it’ll be as fast as all the others in tasks that aren’t thread-heavy workloads.
That, however, doesn’t mean everyone should start saving to buy a $1,000 CPU. No, for most people, the dynamic doesn’t change. For the 80 percent of you who fall into the average Joe or Jane nerd category, a four-core with Hyper-Threading still offers the best bang for the buck. It won’t be as fast as the eight-core, but unless you’re really working your rig for a living, made of money, or hate for your Handbrake encodes to take that extra 25 minutes, you can slum it with the Core i7-4790K chip. You don’t even have to heavily overclock it for the performance to be extremely peppy.
For the remaining 20 percent who actually do a lot of encoding, rendering, professional photo editing, or heavy multi-tasking, the Core i7-5960X stands as the must-have CPU. It’s the chip you’ve been waiting for Intel to release. Just know that at purely stock speeds, you do give up performance to the Core i7-4790K part. But again, the good news is that with minor overclocking tweaks, it’ll be the equal or better of the quad-core chip.
What’s really nice here is that for the first time, Intel is giving its “Extreme” SKU something truly extra for the $999 they spend. Previous Core i7 Extreme parts have always been good overclockers, but a lot of people bypassed them for the midrange chips such as the Core i7-4930K, which gave you the same core counts and overclocking to boot. The only true differentiation Extreme CPU buyers got was bragging rights. With Haswell-E, the Extreme buyers are the only ones with eight-core parts.
Bang-for-the-buck buyers also get a treat from the six-core Core i7-5820K chip. At $389, it’s slightly more expensive than the chip it replaces—the $323 Core i7-4820K—but the extra price nets you two more cores. Yes, you lose PCIe bandwidth but most people probably won’t notice the difference. We didn’t have a Core i7-5820K part to test, but we believe on our testing with the Core i7-5960X that minor overclocking on the cheap Haswell-E would easily make it the equal of Intel’s previous six-core chips that could never be had for less than $580.
And that, of course, brings us to the last point of discussion: Should you upgrade from your Core i7-4960X part? The easy answer is no. In pure CPU-on-CPU showdowns, the Core i7-4960X is about 20 percent slower in multi-threaded tasks, and in light-duty threads it’s about the same, thanks to the clock-speed advantage the Core i7-4960X has. There are two reasons we might want to toss aside the older chip, though. The first is the pathetic SATA 6Gb/s ports, which, frankly, you actually need on a heavy-duty work machine. The second reason would be the folks for whom a 20 percent reduction in rendering time would actually be worth paying for.
Click the next page to check out our Haswell-E benchmarks.
Haswell-E benchmarks overview
We like to give you the goods on a nice table but not everyone is familiar with what we use to test and what exactly the numbers means so let’s break down some of the more significant results for you.
Cinebench 15 single-threaded performance
We used Maxon’s Cinebench 15 benchmark to see just how fast the trio of chips would run this 3D rendering test. Cinebench 15 allows you to restrict it from using all of the cores or just one core. For this test, we wanted to see how the Core i7-5960X “Haswell-E” would do against the others by measuring a single core. The winner here is the Core i7-4790K “Devil’s Canyon” chip. That’s no surprise—it uses the same microarchitecture as the big boy Haswell-E but it has a ton more clock speed on default. The Haswell-E is about 21 percent slower running at 3.5GHz. The Devil’s Canyon part is running about 900MHz faster at 4.4GHz. Remember, on default, the Haswell-E only hits 3.5GHz on single-core loads. The Haswell-E better microarchitecture also loses to the Core i7-4960X “Ivy Bridge-E,” but not by much and that’s with the Ivy Bridge-E’s clock speed advantage of 500MHz. Still, the clear winner in single-threaded performance is the higher-clocked Devil’s Canyon chip.
Winner: Core i7-4790K
Cinebench 15 multi-threaded performance
You don’t buy an eight-core CPU and then throw only single-thread workloads at it, so we took the handcuffs off of Cinebench 15 and let it render with all available threads. On the Haswell-E part, that’s 16 threads of fun, on Ivy Bridge-E it’s 12-threads, and on Devil’s Canyon we’re looking at eight-threads. The winner by a clear margin is the Haswell-E part. Its performance is an astounding 49 percent faster than the Devil’s Canyon and about 22 percent faster than Ivy Bridge-E. We’ll just have to continue to remind you, too: this is with a severe clock penalty. That 49-percent-faster score is with all eight cores running at 3.3GHz vs all four of the Devil’s Canyon cores buzzing along at 4.4GHz. That’s an 1,100MHz clock speed advantage. Ivy Bridge-E also has a nice 700MHz clock advantage than Haswell-E. Chalk this up as a big, huge win for Haswell-E.
Winner: Core i7-5960X
We wanted a second opinion on rendering performance, so we ran POV-Ray, a freeware ray tracer that has roots that reach back to the Amiga. Again, Haswell-E wins big-time with a 47 percent performance advantage over Devil’s Canyon and a 25 percent advantage over Ivy Bridge-E. Yeah, and all that stuff we said about the clock speed advantage the quad-core and six-core had, that applies here, too. Blah, blah, blah.
Winner: Core i7-5960X
Premiere Pro CS6 performance
One sanity check (benchmark results Intel produces to let you know what kind of performance to expect) said Haswell-E would outperform quad-core Intel parts by 45 percent in Premiere Pro Creative Cloud when working with 4K content. Our benchmark, however, doesn’t use 4K content yet, so we wondered if our results would be similar. For our test, we render out a 1080p-resolution file using source material shot by us on a Canon EOS 5D Mk II using multiple timelines and transitions. We restrict it to the CPU rather than using the GPU as well. Our result? The 3.3GHz Haswell-E was about 45 percent faster than the 4.4GHz Devil’s Canyon chip. Bada-bing! The two extra cores also spit out the render about 19 percent faster than the six-core Ivy Bridge-E. That’s fairly consistent performance we’re seeing between the different workload disciplines of 3D rendering and video encoding so far, and again, big, big wins for the Haswell-E part.
Winner: Core i7-5960X
Handbrake Encoding performance
For our encoding test, we took a 1080p-resolution video file and used Handbrake 0.9.9 to transcode it into a file using the Android tablet profile. Handbrake is very multi-threaded and leverages the CPU for its encoding and transcoding. Our results were still fairly stellar, with Haswell-E CPU performing about 38 percent faster than the Devil’s Canyon part. Things were uncomfortably close with the Ivy Bridge-E part though, with the eight-core chip coming in only about 13 percent faster than the six-core chip. Since the Ivy Bridge-E cores are slower than Haswell cores clock-for-clock, we were a bit surprised at how close they were. In the past, we have seen memory bandwidth play a role in encoding, but not necessarily Handbrake. Interestingly, despite locking all three parts down at 2,133MHz, the Ivy Bridge-E does provide more bandwidth than the Haswell-E part. One other thing we should mention: Intel’s “sanity check” numbers to let the media know what to expect for Handbrake performance showed a tremendous advantage for the Haswell-E. Against a Devil’s Canyon chip, Haswell-E was 69 percent faster and 34 percent faster than the Ivy Bridge-E chip. Why the difference? The workload. Intel uses a 4K-resolution file and transcodes it down to 1080p. We haven’t tried it at 4K, but we may, as Intel has provided the 4K-resolution sample files to the media. If true, and we have no reason to doubt it, it’s a good message for those who actually work at Ultra HD resolutions that the eight-cores can pay off. Overall, we’re declaring Haswell-E the winner here.
Winner: Core i7-5960X
X264 HD 5.01 Pass 1 performance
We’ve been using TechArp.com’s X264 HD 5.0.1 benchmark to measure performance on new PCs. The test does two passes using the freeware x264 encoding library. The first pass is seemingly a little more sensitive to clock speeds and memory bandwidth rather than just pure core count. A higher frame rate is better. The first pass isn’t as core-sensitive, and memory bandwidth clock speed have more dividends here. Haswell still gives you a nice 36 percent boost over the Devil’s Canyon but that Ivy Bridge-E chip, despite its older core microarchitecture, comes is only beaten by 12 percent—too close for comfort. Of course, we’d throw in the usual caveat about the very large clock differences between the chips, but we’ve already said that three times. Oh, and yes, we did actually plagiarize by lifting two sentences from a previous CPU review for our description. That’s OK, we gave ourselves permission.
Winner: Core i7-5960X but not by much
X264 HD 5.01 Pass 2 performance
Pass two of the X264 HD 5.01 benchmark is more sensitive to core and thread counts, and we see the Haswell-E come in with a nice 46 percent performance advantage against the Devil’s Canyon chip. The Ivy Bridge-E, though, still represents well. The Haswell-E chip is “only” 22 percent faster than it. Still, this is a solid win for the Haswell-E chip. We also like how we’re seeing very similar scaling in multiple encoding tests of roughly 45 percent. With Intel saying it’s seeing 69 percent in 4K resolution content in Handbrake, we’re wondering if the Haswell-E would offer similar scaling if we just moved all of our tests up to 4K.
Winner: Core i7-5960X
Click the next page for even more Haswell-E benchmarks.
Stitch.EFx 2.0 Performance
Again, we like to mix up our workloads to stress different tasks that aren’t always multi-threaded to take advantage of a 12-core Xeon chip. For this test, we shot about 200 images with a Canon EOS 7D using a GigaPan motorized head. That’s roughly 1.9GB in images to make our gigapixel image using Stitch.EFx 2.0. The first third of the render is single-threaded as it stitches together the images. The final third is multi-threaded as it does the blending, perspective correction, and other intensive image processing. It’s a good blend of single-threaded performance and multi-threaded, but we expected the higher clocked parts to take the lead. No surprise, the Devil’s Canyon 4.4GHz advantage puts it in front, and the Haswell-E comes in about 14 percent slower with its 1.1GHz clock disadvantage. The clock speed advantage of the 4GHz Ivy Bridge-E also pays dividends, and we see the Haswell-E losing by about 10 percent. The good news? A dual-core Pentium K running at 4.7GHz coughed up a score of 1,029 seconds (not represented on the chart) and is roughly 22 percent slower than the CPU that costs about 11 times more.
Winner: Core i7-4790K
The popular and free zip utility, 7-Zip, has a nifty built-in benchmark that tells you the theoretical file-compression performance a CPU. You can pick the workload size and the number of threads. For our test, we maxed it out at 16-threads using an 8MB workload. That gives the Haswell-E familiar advantage in performance—about 45 percent—over the Devil’s Canyon part. Against that Ivy Bridge-E part though, it’s another uncomfortably close one at 8 percent. Still, a win is a win even if we have to say that if you have a shiny Core i7-4960X CPU in your system, you’re still doing fine.
Winner: Core i7-5960X
Sisoft Sandra Memory Bandwidth (GB/s)
Since this is the first time we’re seeing DDR4 in a desktop part, we wanted to see how it stacked up in benchmarks. But, before you get too excited, remember that we set all three systems to 2133 data rates. The Devil’s Canyon part is dual-channel and the Ivy Bridge-E and Haswell-E are both quad-channel. With the memory set at 2133, we expected Haswell-E to be on par with the Ivy Bridge-E chip, but oddly, it was slower, putting out about 40GB/s of bandwidth. It’s still more than the 27GB/s the Devil’s Canyon could hit, but we expected it to be closer to double of what the Ivy Bridge-E was producing. For what it’s worth, we did double-check that we were operating in quad-channel mode and the clock speeds of our DIMMs. It’s possible this may change as the hardware we see becomes more final. We’ll also note that even at the same clock, DDR4 does suffer a latency penalty over DDR3. That would also be missing the point of DDR4, though. The new memory should give us larger modules and hit higher frequencies far easier, too, which will nullify that latency issue. Still, the winner is Ivy Bridge-E.
Winner: Core i7-4960X
3DMark Firestrike Overall Performance
Even though 3DMark Firestrike is primarily a graphics benchmark, not having a 3DMark Firestrike score is like not having coffee in the morning. Basically, it’s a tie between all three chips, and 3DMark Firestrike is working exactly as you expect it to: as a GPU benchmark.
3DMark Firestrike Physics Performance
3DMark does factor in the CPU performance for its physics tests. It’s certainly not weighted for multi-core counts as other tests are, but we see the Haswell-E with a decent 29 percent bump over the Devil’s Canyon chip. But, breathing down the neck of the Haswell-E is the Ivy Bridge-E chip. To us, that’s damned near a tie. Overall, the Haswell-E wins, but in gaming tasks—at stock clocks—paying for an 8-core monster is unnecessary except for those running multi-GPU setups.
Winner: Core i7-5960X
Valve Particle Benchmark Performance
Valve’s Particle test was originally developed to show off quad-core performance to the world. It uses the company’s own physics magic, so it should give some indication of how well a chip will run. We’ve long suspected the test is cache and RAM latency happy. That seems to be backed by the numbers because despite the 1.1GHz advantage the Devil’s Canyon chip has, the Haswell-E is in front to the tune of 15 percent. The Ivy Bridge-E chip though, with its large cache, lower latency DDR3, and assloads of memory bandwidth actually comes out on top by about 3 percent. We’ll again note the Ivy Bridge-E part has a 700MHz advantage, so this is a very nice showing for the Haswell-E part.
Winner: Core i7-4960X
Dirt Showdown low-resolution performance
For our gaming tests, we decided to run the games at 1366x768 resolution and at very low settings to take the graphics card out of the equation. In one way, you imagine this as what it would look like if you had infinitely powerful graphics cards in your system. As most games are not multi-threaded and are perfectly fine with a quad-core with Hyper-Threading, we fully expected the parts with the highest clock speeds to win all of our low-resolution, low-quality tests. No surprise, the Devil’s Canyon part at 4.4GHz private schools the 3.3GHz Haswell-E chip. And, no surprise, the 4GHz Ivy Bridge-E also eats the Haswell-E’s lunch and drinks its milk, too.
Winner: Core i7-4790K
Dirt Showdown 1080p, ultra performance
To make sure we put everything in the right context, we also ran the Dirt Showdown at 1920x1080 resolution at Ultra settings. This puts most of the load on the single GeForce GTX 780 we used for our tests. Interestingly, we saw the Haswell-E with a slight edge over the Devil’s Canyon and Ivy Bridge-E parts. We’re not sure, but we don’t think it’s a very significant difference, but it’s still technically a win for Haswell-E.
Winner: Core i7-5960X
Hitman: Absolution, low quality, low performance
We did the same with Hitman: Absolution, running it at low resolution and its lowest settings. The Haswell-E came in about 12 percent slower the Devil’s Canyon part and 13 percent slower than the Ivy Bridge-E.
Winner: Core i7-4960X
Hitman: Absolution, 1080p, ultra quality
Again, we tick the settings to an actual resolution and quality at which people actually play. Once we do that, the gap closes slightly, with the Haswell-E trailing the Devil’s Canyon by about 8 percent and the Ivy Bridge-E by 9 percent. Still, these are all very playable frame rates and few could tell the difference.
Winner: Core i7-4960X
Tomb Raider, low quality, low resolution.
We did the same low quality, low resolution trick with Tomb Raider and while need to see 500 frames per second, it’s pretty much a wash here.
Tomb Raider, 1080p, Ultimate
At normal resolutions and settings we were a little surprised, as the Haswell-E actually had a 15 percent advantage over the Devil’s Canyon CPU. We’re not exactly sure why, as the only real advantage we can see is memory bandwidth and large caches on the Haswell-E part. We seriously doubt it’s due to the number of CPU cores. The Haswell-E also has a very, very slight lead against the Ivy Bridge-E part, too. That’s not bad considering the clock penalty it’s running at.
Winner: Core i7-5960X
Metro Last Light, low resolution, low quality
In Metro Last light, at low settings it’s a wash between all of them.
Metro Last Light, 1080p, Very High quality
Metro at high-quality settings mirrors that of Hitman: Absolution, and we think favors the parts with higher clock speeds. We should also note that none of the chips with the $500 graphics card could run Metro at 1080p at high-quality settings. That is, of course, you consider 30 to 40 fps to be “smooth.” We don’t. Interestingly, the Core i7-4690X was the overall winner.
Winner: Core i7-4960X
Conclusion: If you skipped to the very last page to read the conclusion, you’re in the wrong place. You need to go back to page 4 to read our conclusions and what you should buy. And no, we didn’t do this to generate just one more click either though that would be very clever of us wouldn’t it?