Read more.Motherboard makers are again using armour/heatsinks that act as heat insulation.
Read more.Motherboard makers are again using armour/heatsinks that act as heat insulation.
Intel never cared for overclockers anyway. Perhaps the board makers do, but this launch was sprung on them and it's not like Skylake X is anything similar to a 7700K. They would much rather have boards that work but are no good for OC, than have no boards and wait for strong OC capability.
Maybe it's just me but it seems a bit whiney and arrogant to be complaining about OC capability on SKL-X.
I'd consider myself the target demographic for these sorts of products, but I absolutely loath the aesthetic and questionable practicality of some features, such as these ridiculous 'heatsinks'.
Does anyone actually like the appearance of these boards?
They pay a premium mostly for the extra cores, (which already run at reasonably high frequency if temperatures and power draw are anything to judge by) not for the ability to tinker. Most of these CPUs come unlocked because for these prices it would be disrespectful of Intel to offer anything else. They're just modified server chips being pushed to start with on motherboards that were not prepared in time. High core parts already come with competitive IPC and clockspeed to Ryzen so overclocking is indeed a luxury that Intel does not have to pander for. I'm sure they would have PREFERRED to but there's no requirement.
It is relevant, yes, but I feel the purpose of the video is supposed to be entirely informative, warning against attempting OC without extensive preparation and give tips on how to work around the lacking performance from the motherboards. Not to take digs at Intel and the manufacturers. Almost anyone who is interested in overclocking an SKL-X or KL-X CPU will have the knowledge that these CPUs came out too fast just to respond to Ryzen and Threadripper and should understand that new tech comes with flaws almost every time.
When Ryzen's boards came out they were riddled with BIOS issues, hampering performance in stock configurations as well as for overclocking. But people didn't complain much about AMD or the boards being bad, they just waited for BIOS revisions and updates from AMD. It was... almost civil.
I got too little sleep last night, sorry.
Would do if they were less.. pointed angles, big blocky heatsinks, and clashing colours. Simpler versions of what they have (preferably with functional heat management) and yes I would.
Sorry, but I don't buy that. That vast majority of consumers don't even need the 8 threads available on Intel's mainstream consumer platform. People who are buying specifically for these kinds of cores counts - 8+ - only really fall into two categories. The first group are people who need high core counts for productivity purposes. The second group is enthusiasts and overclockers.
Of course, for productivity there's the Xeon range, offering the same core counts at lower prices (balanced by lower stock clock speeds, of course) but with better business support and longevity than Intel offers for no-Pro Core products. If you're buying for extra threads for productivity, the only good reasons to go Core X would be if you're home-building (in which case you're probably an enthusiast), or if for some reason you need lots of threads and high clock speeds (a very niche market). That suggests that the majority of people who buy Core X are enthusiasts, and even if overclocking potential isn't the most important factor in their decision it is going to matter to them.
You're right, but this still doesn't change Intel's stance on what these CPUs should do IMO. Unlocking them is a sign that Intel wanted enthusiasts to get them - sure - but why would they rush a release, assume motherboards would be up to scratch and have the chips run relatively high power draw to start with if they were aiming for enthusiasts? They didn't even replace TIM with solder, after continual signs from the enthusiast community that it isn't what they wanted.
If Intel could get away with locking these chips back up they probably would. In stock configuration they provide power above and beyond R7 in most cases and that's the box ticked. Complaining to Intel is as fruitless an endeavor as ever, and motherboard makers will be doing whatever they can to improve anyway having seen the results of the release (I would be surprised if they weren't aware how hard the boards worked for OCs before selling them). So, back to the point - is it whiney or arrogant to complain about lacking OC?
I'm pretty sure you've answered all of your other points with the first one
Without doing something Intel faced trying to launch the latest Core X against an available platform that was already established, had up to 16 cores available, more PCIe lanes, and potentially a better overclocking platform (even if the processors themselves wouldn't clock as high). Oh, and AMD is likely to be significantly cheaper per-core, too.
So they rushed the launch. That means they've got their platform out the door first, while AMD's core count tops out at 8 on a platform that has a lot less PCIe lanes. There's actually reasons beyond "faster in a straight line" to buy Intel now - if you want more than one x16 PCIe slot, or you want more than 2 memory channels, or more than 8 cores. And as a package those are things that some enthusiasts will care about.
Obviously Intel thought that being first to market with Core X (vs Threadripper) was a sufficient advantage to cover any shortcomings in the platform. They clearly made the same assumption you did - that the overclocking features weren't what people cared about. And perhaps they're actually right - we don't know how many Core X systems/processors have been bought, and how many people have tried to overclock them, or even how many people have been put off by thew platform's poor design for overclocking.
But to bring this back to your original point:
Intel's branding says otherwise. They refer to the top end processors as "Extreme Edition". They make a bullet point of them being "fully unlocked for performance tuning". And in the Computex announcement: "we’re introducing the new Intel® x299 chipset, which adds even more I/O and overclocking capabilities.".. Maybe it's just me but it seems a bit whiney and arrogant to be complaining about OC capability on SKL-X. ...
Well, guess what? Tell someone you're adding more overclocking capabilities, and they're going to take you to task if your platform isn't OC friendly. So on the whole, I think it's entirely justified to complaining about Skylake-X's OC performance. It's one of the things Intel were trying to sell it on.
So we've got the triumvirate of insulation? Insulating heatspreaders, insulating flaps to cover m.2 SSDs, and now insulating 'heatsinks' on the VRMs?
His criticism of the cable temperature is a little harsh, I feel - the temp at the connector will be higher due to the higher resistance at the connections, as well as the insulating effect of the plastic connector
Ryzen was rushed, but it also came with half the price tag. VRM heatsink design isn't that hard, people have been tinkering with it for years - making that kind of rookie mistake when they're charging so much for a motherboard is silly
I am very surprised hexus missed that on its review...
https://www.newegg.com/Product/Produ...82E16819117795
"Unlocked Processor "
Not arrogant or whiney to complain about NOT getting what was ADVERTISED. From the newegg link above, Intel's product info:
"Overclocking: Gain freedom to push the limits exactly how and where you want with this highly overclockable processor.1 Precisely tune voltage per core, and make exact adjustments with enhanced memory overclocking. It's another way to custom-fit your processor to your exact specifications."
Those are INTEL WORDS. Umm, there is a whole section devoted to the advertising on newegg's page. It seems the statement above, at the very least, needs some work huh? Is it arrogant to sell me something for a grand that can't do what it says? The boards advertise this same crap too so again...Why are you blaming the customer for expecting what they paid for in BOTH cases? Board and CPU should OC here. This is the top end you can buy, failure here is unacceptable IMHO. When the top OC'er in the world says all your boards are designed wrong, there is a problem...ROFL. I'm sure you're right about Intel wanting all chips locked. But that isn't what this is about. This is about saying the most expensive boards should OC, and the best chip to do it with is Intel's new Skylake-X UNLOCKED chips, but then having the combo's results fizzle in practice. Oh, and the pain of realizing this AFTER you've paid for both...LOL.
Maybe you think it's crazy, but I expect products to do what they say when sold to me. It's an even larger problem when one of the major features is essentially broken. It doesn't matter what you think Intel thinks about OCing here. What matters is they are selling these chips as GREAT overclockers and so are board makers with regards to their part of the combo.
https://www.intel.com/content/www/us.../i9-7900x.html
Intel's own page stating you can "OVERCLOCK WITH CONFIDENCE".
OK. I guess I should expect good overclocking correct?
https://www.intel.com/content/www/us...rocessors.html
Maybe you shouldn't dedicate whole sections of your site to it...LOL. Whiney? LOL. It's arrogant to post pages like this, and advertise like they did for the current cpus, but fail to deliver on such features miserably is it not? Whatever, I digress.
Last edited by peterb; 30-06-2017 at 08:53 AM. Reason: Merge consecutive posts
This has been picked up by one of the world's top overclockers.. It's a little unfair to expect a company like Hexus to pick up on this; they're providing reviews to the masses, based on the platform as a whole, whereas De8auer is addressing those concerned by the bleeding edge of performance, where a couple of degrees can mean the difference between 4.9GHz and 5.0+GHz.
Still, a very interesting video. I really hope the problems aren't purely caused by people who expect a stylish motherboard to compliment their RGB RAM, all of which is visible through their windowed case.
There's a lot of air at the connection junction perhaps? Or maybe the pins aren't well connected to the wires.
And yes, they should be.. but that doesn't stop either a) people breaking PCI-E spec and pushing more Amperage down the cable than rated and b) cables/connectors not actually meeting their ratings.
The issue isn't the resistance of each half of the connector, it's the point where they touch each other. The connection between the two parts is imperfect and has a much higher resistance than any other part of the circuit.
As to component ratings, they're generalised to "standard" environments, and specific to the individual component. They take no account of the specific environment, or of other components in close proximity. So the fact that a single connector is operating within its rating doesn't mean the entire system is operating safely - the build up of heat could impact other components.
There are currently 1 users browsing this thread. (0 members and 1 guests)