ASRock turns to an external clock generator for BCLK overclocking

ASRock just won't do what Intel tells it to. The company may have issued BIOS updates that removed its SkyOC base-clock (or BCLK) overclocking features from many of its motherboards, but it's not abandoning the locked-CPU overclocking dream just yet. This time around, the company is taking the battle to the hardware. Its Fatal1ty H170 Performance/Hyper and Fata1ty B150 Gaming K4/Hyper boards both rely on an external clock generator for the BCLK signal. This generator appears to allow for base-clock overclocking of locked chips on these motherboards.

ASRock is rather taciturn about the purpose of this clock generator on its product pages for the Hyper boards. It describes the feature as an "external base clock generator which provides a wider range of frequencies and more precise clock waveforms." The spec sheet and manual both refer to "ASRock BCLK Full-range Overclocking," however, so it seems the base-clock overclocking game is back on for now.

Neither board appears to be for sale yet, but the non-Hyper B150 Gaming K4 is only $100 at Newegg, and the H170 Performance is $112. We'd expect prices for the Hyper boards to slot in near those figures.

Comments closed
    • bfar
    • 4 years ago

    Why not z170?

    Do Asrock’s B150 and H170 have the power delivery circuitry to handle good bclk overclocks?

    • kuttan
    • 4 years ago

    ASRockz!!

      • w76
      • 4 years ago

      Need more fiber.

      • Neutronbeam
      • 4 years ago

      You bet your AS!

    • iatacs19
    • 4 years ago

    Bad ass! I like!

    • UnfriendlyFire
    • 4 years ago

    Intel: “Goddamit ASRock, you’re eating into our profit margins!”

      • Krogoth
      • 4 years ago

      Not really.

      The main concern with overclocking is that shady OEMs can use it as a means to upsell silicon binned silicon. It was reason why locked chips were implemented in the first place.

        • UnfriendlyFire
        • 4 years ago

        I’m sure Intel’s lawyers would have a field day with such OEMs.

        I’m also certain that AMD would back away from an OEM that was caught fudging with Intel’s CPUs. Say, when was the last time a VIA CPU was on a desktop board?

          • Krogoth
          • 4 years ago

          That’s exactly happened during the 1990s. There were shady OEMs that were selling systems equipped with a “Pentium 90” when in practice it was really a Pentium 60 overclocked to Pentium 90.

          After that debacle is when Intel and later AMD decided to lock down the multiplier on their CPUs.

          Intel and AMD never really care about enthusiasts that overclocking their stuff since it voids the product warranty on CPUs and they usually know when a chip died from abuse when it gets back to the fab.

          It is only a recent development that overclocking is permitted on certain models (K series and X series) as long as you didn’t kill the chip by openly abusing it (ie lapping it, removing IHS, extreme overvolting etc).

          To be honest though, it is really difficult to kill a CPU with arm-chair overclocking. You really have to pump a ton of volts and/or expose the core (remove the IHS) to physical damage from HSF.

        • w76
        • 4 years ago

        Frankly, I don’t believe that for a second, and if you do you’ve drank the corporate Kool-aid. It’s all about market segmentation. That makes huge logical sense. Supposed phantom OEMs with minuscule market shares being shady doesn’t seem like a reason, otherwise, to deny enthusiast consumers of your product what they want. Capturing more surplus/utility/profit, though, through discrimination/segmentation is economics 101. Even movie theaters do it.

          • Krogoth
          • 4 years ago

          No it is what actually happened.

          Intel and AMD really do not care about overclocking crowd as far as them affecting the bottom line. For them the entire affair is at user’s own risk and expense. The product warranty makes it very clear. Overclocking the “K” and “X” series is basically we will cover it as long as you don’t openly abuse the product.

          It is a completely different story for OEMs. Not only it hurts customers who don’t know any better. It damages Intel and AMD’s reputation in the eyes of businesses and mainstream customers who end-up blaming their instabilities with a system that had a clandestine overclock on the platform rather then OEM (Just look what happen with P-III 1.13Ghz debacle). It also happens to cut into the bottom line since OEMs are going try to upsell binned silicon. Intel and AMD sees none of that.

            • w76
            • 4 years ago

            I still don’t buy that, they can handle OEMs by way of contractual agreements, except for the very smallest, who have minuscule market share in 2016. It also makes little sense for retail box kits; consumers sophisticated enough to buy a part to assemble on their own are presumably savvy enough to manage the decision to over clock. (But why allow that when you can force them in to a higher-cost 6500K or 6700K?)

            It’s like a couple weeks ago, when the word came down from upper management at corporate HQ, over time in production had to go. Because, it might lead to injury. Yeah. Totally not that I can see myself the OT hurts the bottom line. Not that at all! Totally because Jimmy lost a finger in ’99 by bypassing the saw safeties.

            Next you’ll tell me the 6700K doesn’t have vPro but the 6700 does because, I don’t know, some odd answer that’s not naked price discrimination.

            Now, you did say “we,” if you work for Intel then I accept you have to parrot the corporate line. But I don’t have to believe it’s any more sincere than, say, the copy-and-paste corporate statements in support of gender equality or pick-your-CSR-issue.

          • flip-mode
          • 4 years ago

          Kroggy isn’t telling it quite right. OEMs didn’t sell “upbinned silicon”. They sold whole PCs with the CPU overclocked to a higher spec and labeled the PC as a higher spec’d computer.

            • ImSpartacus
            • 4 years ago

            I think what he means that the oems binned cheaper parts and found the ones that could semi-reliably overclock to whatever spec they wanted. The remainder would be sold in a stock machine.

            I don’t remember this being much of a serious issue, but it’s been a while and I probably forgot.

            • w76
            • 4 years ago

            That supports my point. We all kinda remember rumors of such from the 90s, but it’s been many years and many OC-friendly parts since then. It just doesn’t stand up to basic logic; what’s the more likely motivator for a decision, the aspect where it keeps people from paying less for the same performance and probably lowers total volume a bit, or the aspect where the people who still buy are paying more money?

            Also, if people don’t think stunts like Devils Canyon and more specifically that Anniversary Edition OCing beast weren’t market experiments, well… Intel almost certainly had a dedicated in-house economist studying how those parts influenced the market; prices, volumes of different chips, purchasing decisions.

    • ronch
    • 4 years ago

    It’s been nice knowing you, AsRock. ☺

    • chuckula
    • 4 years ago

    Aside from the fact that it’s quite easy to find firmware versions that still allow for BCLK overclocking, this once again goes to show that Intel really didn’t go out of its way to prevent overclocking of locked Sklyake parts.

    One more thing: These boards not only enable BCLK overclocking, but you’ll note that neither one uses the high-end Z170 chipset. So not only did Intel not really restrict BCLK overclocking with Skylake, but overclocking is now available on a much larger range of platforms than it was in the past.

    Go ahead and downthumb all day, I’d rather have an overclocked i3 than any FX chip that AMD will actually have for sale in 2016.

      • slowriot
      • 4 years ago

      Your definition of “really didn’t go out of its way” for Intel here is bizarre. They pressured OEMs to the point that only one, ASRock, is even trying to offer BCLK overclocking now. And these boards will carry a further premium over their standard versions. Pushing the motherboard prices into spaces occupied by a number of good Z170 options.

      At this point its really hard to see being limited to two motherboards from the same manufacturer to save $50-$70 for a theoretical overclock goal as a “win” for the consumer.

      But I guess since Intel doesn’t have the magic ability to purge those BIOS from the Internet or hasn’t [b<]yet[/b<] done something about this is Intel "not really going out of its way."

        • maxxcool
        • 4 years ago

        Intel could have EASILY instituted a check to see the frequency of various sections of the cpu and shut it down or down-clock via internal multipliers.

        But I DO expect in the future that INTEL *will* frequency lock cpus.

          • slowriot
          • 4 years ago

          “Easy” for Intel has nothing to do with it. Of course stopping the practice out right was easy, but they’re trying to play both sides of the game here. Theoretically they’ve allowed BCLK overclocking to continue but only if you jump thru a set of hoops…. like either running your motherboard on an out-of-date and/or unsupported BIOS level or buying one of these ASRock models which will cost as much as some of their own Z170 models (let alone other competitors). This move would hypothetically give them the PR room to say “We didn’t stop it” but in reality have putted enough hassle to make a lot of people second guess the decision of buying into it.

          Plus there was a brief period where it looked like Intel wasn’t going to slow BCLK overclocking at all and then they took action. Who knows what will happen here. I hope for the sake of whoever buys these boards nothing, but I personally wouldn’t be willing to save only $50-$70 and bet on that (since now the only money you’re saving for a “supported” solution is on the CPU).

          Intel doesn’t care about AMD, at least not right now. I imagine they’re far more concerned about getting people concerned about performance who are upgrading now to do so with a move to a higher end i5 than an a OC’d i3.

          • bhtooefr
          • 4 years ago

          Watch Kaby Lake or maybe Cannonlake have the reference oscillator under the IHS, instead of on the motherboard. (And, really, there’s no reason they can’t do it on KBL – oscillator on package, disconnect the CLK input pin from the bottom side, and that’s it, no more BCLK overclocking.)

          (That’s really what ASRock is doing, they’re just changing the reference oscillator frequency to affect BCLK – basically, this is how you overclocked FSB before BCLK could be manipulated in the CPU directly.)

          Of course, the trick with raising reference clocks is that [i<]everything[/i<] is then overclocked... so detecting it is actually harder than it seems. You'd have to go to the real-time clock (which has its own reference clock), and that isn't something that microcode should really be doing.

        • chuckula
        • 4 years ago

        What premium are you talking about exactly. The article didn’t publish their prices.

        Even if there is a “premium” then so the hell what. Last time I checked these motherboard makers were still pushing out overpriced premium boards for geriatric AM3+ parts even in 2016. I didn’t see you getting all high & mighty about that.

          • slowriot
          • 4 years ago

          You think ASRock won’t charge more for these over the standard models? It’s an assume premium. It might be only $10-$20 but that’s enough to start making the value play questionable on what are already pricey boards for their respective chipsets.

          Intel is acting to protect their own higher profit margin parts, not to squeeze AMD. Who cares about an AM3+ board right now? No one, well except you and only so you can mock it.

        • ImSpartacus
        • 4 years ago

        Save $50-70?

        Bro, 6100s regularly go for $130 ([url=http://pcpartpicker.com/part/intel-cpu-bx80662i36100<]$122 at the time of this post[/url<]). Meanwhile 6600Ks haven't dipped under $200 and usually go for closer to $250 ([url=http://pcpartpicker.com/part/intel-cpu-bx80662i36100<]$245 at the time of this post[/url<]). So it's more like a savings of $110-120. And you don't need to spend all of that on a cooler. If I was in the market, I'd go that route, get an affordable cooler and aim for a couple tenths of a GHz. If I can get above 4.0GHz at all, then that's a win in my book.

          • slowriot
          • 4 years ago

          Why are you jumping all the way from a 6100S to a 6600K? I would rather land somewhere in the middle.

          Frankly, I would rather get a motherboard which has features not focused on overclocking, which tend to be cheaper, and get something like a i5 6400 or 6500. The additional cores are going to be more important than the frequency over the 2-4 years the person buying wants to use this as their gaming rig CPU. Or if you’re not gaming even more important already.

            • ImSpartacus
            • 4 years ago

            I thought you meant replacing a 6100 with a legitimate overclocker, my mistake. However, I wouldn’t consider the stock 6400 or 6500 to be an equivalent replacement to a moderately overclocked 6100 (not 6100S).

            I get that you’re acknowledging that fact, at least for the present (and present benchmarks support it as well).

            However, even in the near future, I’m not convinced that low-clocked many-cored CPUs will be the predominant gaming choice.

            We heard the exact same story close to a decade ago. It was back when the Penryn E8400 was priced similarly to the Conroe Q6600 (G0, of course). Both had the same max multiplier. The dual core E8400 could flirt with nearly 4 GHz while the quad core G0 Q6600 could usually get up to 3 GHz. People told gamers the exact same story that you’re telling right now. They said the Q6600 was an investment and it would be awesome in games in just a couple years, blah blah blah. It didn’t work that way.

            Nine years later, you’re trying to tell that exact same story. And strangely enough, the situation is similar. A cheaper 6100 can pretty routinely get to 4 GHz and often a hair higher without ridiculous voltages. Meanwhile, the stock 6400 and 6500 are locked at 2.7 Ghz and 3.2 GHz in 4-core mode. They gain a couple tenths of a GHz in dual/single-core turbo, but nothing even close to 4 GHz (and that shouldn’t matter since you’re suggesting that their four cores would be utilized).

            I very well could be jaded, but I’ll believe it when I see it. That was my stance nine years ago and it’s my stance today.

            The 6100 is the better choice for today’s gaming situation. Worst case, you’d be out barely a hundred bucks if you decided to upgrade to a beefy kaby lake quad in a year (potentially with a larger gpu & L4, which you can’t get today).

            • MrJP
            • 4 years ago

            I disagree about the E8400 history lesson. I bought one of these at the time when the received wisdom was that a fast dual-core was a better choice for gaming. I later found this sometimes struggled with low minimum frame rates no matter how much I overclocked it, particularly with GTA IV. I upgraded to a Q9550 and this was a big step up. Maximum and average frame rates were similar (predeominantly limited by the graphics card) but minimum rates improved dramatically and the whole experience was much smoother.

            Perhaps this is a corner case with a slightly-shoddily coded console port, but I’m still gaming today with the Q9550 (after upgrading the GPU a few years ago) and this will tide me over until I upgrade the whole system when the new GPUs come out later in the year. I don’t think this would have been the case if I’d stuck with the E8400.

            Based on this experience my next system may well be i7 rather than i5 despite the current received wisdom.

            • ImSpartacus
            • 4 years ago

            I appreciate the anecdote. I’ve heard many similar stories.

            The reoccurring problem is that I haven’t seen extensive reviews that properly explore the topic and come to that conclusion. As you said, your anecdote could possibly be some kind of complex corner case (neither of us can know for sure). We all generally agree to trust reviewers because they bust their ass testing things with the goal of being able to draw actual conclusions. They might find corner cases or other anomalies, but they have enough data to overcome them.

            From the limited (and always growing) pool of data that I’ve found, if you’re on any remotely restrictive budget, you should seriously consider options like a 6100 (overclocked or not).

            If you’re not particularly budget-limited, then you can trivially just pick the “i7”, as you mentioned. There really isn’t any need for “strategy” if you have the means to simply get the best stuff available (or nearly so).

            • MrJP
            • 4 years ago

            It’s not a trivial choice at all, but I am probably thinking longer-term than most given my current system has lasted nearly 8 years now. In the context of the total system cost, I regretted not spending the extra to go higher up the range from the outset last time and it ended up costing more overall.

            Others may be happier to get more bang for their buck in the short term but possibly have to upgrade again sooner. Different strokes etc, though personally I think dropping below 4 physical cores these days may be a step too far.

        • DoomGuy64
        • 4 years ago

        That, and didn’t they issue some sort of microcode update directly over windows, not to mention bclk overclocking disables certain extensions like avx?

        • MOSFET
        • 4 years ago

        I think ASRock printed BLCK-overclocking on so many (relatively highly-priced) motherboard boxes they pretty much had to find a way to keep it. Fury of Intel vs fury of consumers…

      • maxxcool
      • 4 years ago

      🙂 5ghz air cooled i3 taking a belt to EVERY current FM-socket AMD cpu in +discrete gaming. WIN!

        • slowriot
        • 4 years ago

        Which CPU cooler? Which motherboard? Voltage? Any mods like delidding?

      • Goty
      • 4 years ago

      Yeah, Intel didn’t go out of its way to restrict BCLK overclocking just like NVIDIA didn’t go out of its way to restrict heterogeneous PhysX setups.

      There’s a difference between not supporting a feature and explicitly disabling it. This is simply a workaround that Intel doesn’t have control over.

      • Freon
      • 4 years ago

      [url<]https://www.youtube.com/watch?v=LQCU36pkH7c[/url<]

      • kuttan
      • 4 years ago

      You need to use an old BIOS till the end of your PC life. Yup Intel wont prevent you from using an old BIOS as long as you want. If Intel was so friendly towards Overclocking there CPUs then they should have enabled BCLK overclock options on all of their motherboard chipsets and CPUs just like in the Nehalem days were you can overclock any Intel Corei5/i7 CPU on any motherboard chipset by way of BCLK overclock. But due to Intels greed they blocked every possible overclock with their locked CPUs and wanted consumers to buy their top end chipset motherboard + unlocked CPU combo for overclocking.

      It is really thankful to AsRock and similar motherboard vendors for their sincere attempt to bypass Intels overclock block implementations on their CPUs all for the consumers advantage. You sounds more like a Intel Shareholder than a consumer.

        • UnfriendlyFire
        • 4 years ago

        Intel could persuade MS to drop Windows 10 support for the old BIOS versions… I mean MS intends on not supporting newer hardware (ex: Kaby Lake and Zen) for Windows 7 and 8.

Pin It on Pinterest

Share This