Quad-channel RAM vs. dual-channel RAM: The shocking truth about their performance - williamsherat1979
Corsair
One of the check-sour features of Intel's bragging Haswell-E CPU is support for quad-channel DDR4 memory, but my testing shows it may not matter much.
Think of retentivity channels as shotgun barrels. You know from video games that two barrels are better than one. Now think of quad-channel RAM is the four-barrel scattergun of computers: The to a greater extent memory channels, the more memory bandwidth getable to the CPU.
For each conduct in a modern PC, you need an individual stick of Aries. This also depends, of course, on the Mainframe. Consumer chips such as the Core i7-4790K and the new Pith i7-6700K support upbound to two channels, patc consumer chips so much as the Core i7-5960X can support up to four channels
Normally this doesn't matter. You wouldn't corrupt an costly X99 motherboard and pricy Core i7-5960X, then intentionally gimpiness its quad-channel retentivity performance by installing only two pieces of RAM instead of four.
Asrock Asrock's X99 MiniITX board simply does away with two memory channels to economise space.
Assemble Asrock's X99E-ITX/ac
The trouble? You can't actually in condition Intel's Haswell-E chip and the four retentivity slots it necessarily into a smaller Mini-ITX motherboard. They just don't physically fit victimisation to the full-sized memory modules. Without get at to Intel's Haswell-E CPU, that means miniature PCs are limited to quad-core CPUs at best.
Asrock's crazy solvent was simply to give off 2 of the slots on its X99E-ITX/ac motherboard. Yes, that cuts your bandwidth in uncomplete, but it lets you build such unhinged machines American Samoa this Falcon Northwest Beaver State this exotic CyberPower Trinity Xtreme and run more than four CPU cores. The big question is: How much of a gain behave you take?
How we tested
I decided to test just how much real performance you give up by leaving half your system bandwidth behind, As the Asrock X99E-ITX/Ac is for good limited to dual-conduct memory, the alone way I could test this was to use a life-size X99 motherboard.
For that I turned to a MicroExpress B20 organization we reviewed. It has a full-size Asus X99 In favor of motherboard and a sixer-core Core i7-5820K CPU, along with a GeForce GTX 970 card and 16GB of DDR4/2666 RAM in musculus quadriceps femoris-groove mode, using four 4GB modules. I ran several benchmarks with it in quad-channel mode, then swapped out the four sticks of RAM for two sticks of 8GB DDR4/2666 in dual-channel mode.
I could have just pulled two of the systems' original four memory sticks but I distinct some would be implicated the 16GB vs. 8GB of total RAM would dissemble the results. It wouldn't, but I'll liquid body substance you. So for the record: We're testing 16GB of DDR4/2666 in dual-channel mode vs. 16GB of DDR4/2666 in quad-channel mode.
I simluated dual-channel performance using this MicroExpress B20 scheme with its X99 chipset and Core i7-5820K chip.
Sisoft Sandra Computer storage Bandwidth
My firstborn test was SiSoft Sandra's memory bandwidth test. This jack-of-each trades benchmark suite measures and pokes just about everything in your Microcomputer. Information technology's long been a standard to measure available memory bandwidth in a PC. The results were as expected (and besides a good way to double-check that I hadn't put the modules in the wrong slots). Going from dual-channel DDR4/2666 to musculus quadriceps femoris-channel DDR4/2666 nearly doubles the disposable store bandwidth. Woohoo! Go home, right?
Nope. This graph is likely the only good word for quad-channel memory, but I'll let you relish in the bandwidth for now. Read along for the real performance impact.
Sorry, treble-channel Aries the Ram: Quad-channel is way fitter.
Encoding functioning
Semisynthetic tests measuring the theoretical performance is one thing, but just where does it show up in real tasks? To find kayoed, the next test I threw at the organisation was Handbrake. A popular and rid TV encoder, it's a C.P.U.-heavy test. Equally video encoding is something that's believed to be bandwidth-sensitive, I thought doubling the memory bandwidth would salary off hulky-time. Unfortunately, if you look at the chart below, I proverb zippo. I was quite openmouthed, as I've monthlong believed computer storage bandwidth helps encryption performance. I've really seen it in the other on older hardware platforms, to a fault, and so this was a shocker. I will say: This isn't the modishness, as antithetic encoders and different encryption loads could party favor the increased bandwidth. But today, I'm pretty disappointed.
I sawing machine well-nigh no difference in our encoding try out betwixt multiple-channel and musculus quadriceps femoris-communication channel RAM.
PCMark 8 Creative
My side by side task was PCMark 8's Fanciful Conventional test. This counterfeit test attempts to simulate a workload of photo editing, video encryption, alight gaming and browsing. I running game the conventional luck rather than the GPU dowery to keep the workload closed to the Central processing unit itself. The result was, once more, bad surprising and dissatisfactory.
PCMark 8's Creative Conventional test shows minimal impact too.
PCMark 8 Home
I as wel ran PCMark 8's Home and Work Conventional tasks to interchange awake the workload. Again, nearly double the system memory bandwidth made no more difference. I'm non even active to bother wasting Internet bandwidth with the chart of PCMark 8 Work's result, because it's the same.
PCMark 8 Nursing home Conventional test too shows a tie between dual-channel and quad-channel memory.
WinRAR
Like video encoding, file compression is one of the tasks that typically benefits from boatloads of memory bandwidth. To find proscribed, I reached for WinRAR 5.21 and used its built-in compressing benchmark. Ultimately I saw the enlarged memory bandwidth paying inactive—but not by much.
Contempt file compression normally benefitting from memory bandwith, the difference in WinRAR is pretty tiny.
7Zip
I also unemployed up the beta version of 7Zip and ran its home benchmark. I actually saw a decent boost from 7Zip, but again, I really unsurprising Sir Thomas More. I was almost make to hang information technology up but decided to run some play tests, besides, so keep meter reading.
We come picture an improvement from 7Zip's internal benchamrk but non as a lot as I'd expected.
Tomb Raider
For the final set of tests, I fired astir a few games that aren't on the up-to-date of graphics. The idea is to use older games that would not be bottlenecked by the GeForce GTX 970 in my testbed. Information technology's a good card and a hell of a deal, but it's no Titan X. To remove any nontextual matter chokepoint, I likewise ran the games at a fairly low resolution of 1920×1080, and picked lower image quality settings.
The result? Yup. You guessed it: No diff. Not at supreme or high. Just squint your eyes and venture the results are for some.
Running Grave Raider with the system set to either quad-channel seems to pee no difference either.
BioShock Countless
And yes, BioShock Infinite didn't care either, even when pushing 200 fps. The like Grave Raider, I actually ran the examine at high and average settings but decided not to knock off bandwidth since it didn't matter to.
BioShock Infinite at medium settings also shows very minute performance improvement.
Dirt Showdown
And yes, more of the same in Dirt Showdown. There's no reason to picture you the three otherwise settings I ran, because they're whol just the equivalent. Read on for my conclusion.
Unimproved Face-of also shows it's a yawner
Conclusion
I've written about Asrock's interesting decisiveness to forfeiture memory bandwidth for core count a few times now. The reaction is normally to recoil from those who just don't want to make that compromise in memory bandwidth. Give leading half your retention bandwidth evenhanded to make a smaller system with 6 or eight cores? Never!
I had the same response myself originally. Later on running my tests though, I'm not certainly it matters. I'm in for that someplace out there beneath the pale moonlight, there's a job operating room benchmark that truly pays the dividends you'd expect by doubling the available organization bandwidth, but I'm not seeing it here.
Why? I suspect extraordinary reason might be the massive 15MB cache in the 6-core Burden i7-5820K processor I exploited. The quad-core Core i7-4790K has an 8MB cache. That's virtually dual the cache with only two more than cores added to the equation. Could going to an 8-core Core i7-5960X show up the weaknesses of cutting system memory bandwidth in incomplete? After my tests today, I'm not indeed sure information technology will.
I will enunciat, if I built or bought a full-size X99 Haswell-E machine, I'd still want quad-channel memory, because in that respect's just no reason to open it leading. But if I had to prefer a small box where I got six cores alternatively of 4, and my workloads benefited from the extra CPU cores? I'd have absolutely no trouble qualification that decision to throw memory bandwidth overboard.
Asrock Sacrificing computer memory bandwidth for core count doesn't come out to wounded.
Note: When you buy out something afterwards clicking links in our articles, we may clear a small commission. Read our affiliate link insurance policy for more details.
One of institution fathers of hard-core tech reporting, Gordon has been covering PCs and components since 1998.
Source: https://www.pcworld.com/article/423521/quad-channel-ram-vs-dual-channel-ram-the-shocking-truth-about-their-performance.html
Posted by: williamsherat1979.blogspot.com
0 Response to "Quad-channel RAM vs. dual-channel RAM: The shocking truth about their performance - williamsherat1979"
Post a Comment