Skip to main content

Trick or Treat? CPU Cherry-Picker Silicon Lottery to Close October 31st

CPU stock image
(Image credit: Shutterstock)

Silicon Lottery, the go-to service for professional delidding and CPU "golden sample" acquisition, has announced that it will shutter its business by the end of next month. After seven years of offering its unique services to the world, Silicon Lottery cites reductions in CPU overclocking headroom, the increasing use of thermal solder as a standard option, and market changes as causes for its decision to close the store. The closure will be effective on October 31st, with clients that want to have their CPUs delidded having until November 30th to send in their silicon for processing. That means that if you want to add some extra juice to one of the best CPUs for gaming, you'll have to act fast. 

Overclocking headroom reductions for the latest families of processors from both AMD and Intel appears to be one of the main causes behind Silicon Lottery's closing. Both companies have aggressively improved their boosting algorithms to allow for higher sustained and burst clockspeeds - and these algorithms have become so good that it's now even possible to lose single-threaded performance by overclocking (single-core boosts often reach higher clocks than a stable all-core overclock would). 

At the same time, both AMD and Intel have been increasing the base frequency on their products to offer better "out of the box" performance against their competitors — this also cuts into the overall overclocking headroom.

Another element is that manufacturers themselves (and especially Intel) have been aggressively resorting to binning as a way to expand their product portfolio — Silicon Lottery cites the case for the 11900K essentially being a binned 11700K, which means that there are diminishing returns in selecting through the already-binned 11900K CPUs for any tangible benefits from overclocking.

Another proverbial nail in Silicon Lottery's business coffin is that both AMD and Intel are now employing solder-based TIM (Thermal Interface Material) on their CPUs, which not only reduces the thermal improvement that can be gained from the delidding process, but also makes it a higher-risk process — CPUs have been ripped apart alongside their heatsinks by less cautious delidders. 

Of course, businesses are born out of opportunity. If today the sensible business decision for the company is to shutter their services, maybe tomorrow the market and technological fluctuations will make sense for it to rise again. The administrators at least seem to be holding on to that prospect. Until they're back in business, however, do take a look at our guide for the current best CPUs for gaming or workstation scenarios — we, too, help you do some pre-buy binning on your decisions.

  • watzupken
    I actually never believed that the business is sustainable in the first place. How many people are computer enthusiasts that will want to shell out a significant premium for the absolute fastest chip they can get? The premium for top grade chip is very steep, and generally are of interest only to extreme overclockers, which makes sense for competition. In our day to day usage, that extra 100 to 200 Mhz that we can squeeze out of it is not going to make a material impact to my user experience. In addition, to squeeze out that extra Mhz, you need to maintain some cutting edge cooling solution that can keep up with the heat and power consumption.

    In addition, CPU refresh cycle is quite fast. So after testing and having a bunch of CPUs, it may be hard to sell them off as it gets near to the next cycle. So I do wonder if they managed to sell off all their CPUs in the end without making a loss on any of them to begin with.
    Reply
  • vinay2070
    watzupken said:
    I actually never believed that the business is sustainable in the first place. How many people are computer enthusiasts that will want to shell out a significant premium for the absolute fastest chip they can get? The premium for top grade chip is very steep, and generally are of interest only to extreme overclockers, which makes sense for competition. In our day to day usage, that extra 100 to 200 Mhz that we can squeeze out of it is not going to make a material impact to my user experience. In addition, to squeeze out that extra Mhz, you need to maintain some cutting edge cooling solution that can keep up with the heat and power consumption.

    In addition, CPU refresh cycle is quite fast. So after testing and having a bunch of CPUs, it may be hard to sell them off as it gets near to the next cycle. So I do wonder if they managed to sell off all their CPUs in the end without making a loss on any of them to begin with.
    Most higher end tiers were always sold out in a few days of release. I think they did well until companies started binning it themselves. Ryzen barely had headroom to OC and intel started the same with their later series.
    Reply
  • jp7189
    While I never actually purchased anything from siliconlottery I found their statistics page to be invaluable. I often used it to help determine which processor to buy. AFAIK nowhere else is there a compilation like that.
    Reply
  • Phaaze88
    That's unfortunate.
    Hopefully this helps drive home the fact that the bar for overclocking modern cpus and gpus has been set way too high for most... but it probably won't, since Silicon Lottery(company) wasn't that well known.
    If you don't at least LN2 the things, you're not going to get much out of either.
    The engineers took the fun out of it for most users.
    Intel: already have aggressive turbo boosts out the gate. Nothing like what it used to be in the Sandy Bridge days.
    AMD Ryzen: aggressive turbo boosts and boost shut off point, especially with Ryzen 5000.
    Nvidia: the sneaky Gpu Boost algorithm. So many users out there are clueless on just how 'stable' their OCs really are...
    AMD Radeon: I believe they have something similar to Gpu Boost. Haven't really looked into it.
    Reply