Massman

I was told they were going to test the CPUs up front. But I don't know the procedure, or testing depth, or any more details than that.

Fixed. You don't submit straight on the competition page? I don't understand why this is failing for you, but working for other people.

Might be related to architecture or HTT spec. Sempron is still K10. I've tried multiple boards, multiple BIOSes, multiple CPUs, multiple memory kits, and they all have the exact same walls

Hoping to see more activity, though

Helium takes this below 9 seconds ?

This issue only persists when the base clock frequency is adjusted at runtime. Overall CPU frequency or multiplier adjustments are not causing this. All automatic overclocking is done via the CPU ratio. In other words: except for those who downclock the bclk at runtime, this is a non-issue. That would be 99,999% of the people. I have not seen a single press release where this bug was exploited for marketing purposes. It is very fair to assume this is just an overlooked scenario when making design decisions.

Wow, things are going fast!

The UD7 is a little tricky like that. I have tested 3 different boards, and they all behave the same. At least with the latest non-UEFI BIOS. Air: 275 max stable, sometimes doesn't boot, sometimes loses channel LN2: 335 max stable, after that freeze or no post To get to 300HTT+, you need to dial in the memory subtimings manually. Once you figured out a right problem, save it. Also, you will never be able to boot at the top HTT frequencies. My procedure usually is BOOT 240 -> 275 -> 295 -> 300 (-> 315 (if I want to get close to 330 HTT) This has worked on all three boards. Fyi, the UD3 can do 300HTT on air, and 350HTT on LN2. Less layer seems better for signaling ...

Odd choice of memory divider. In my testing, 3:10 > 1:4 > 3:14 for memory performance (http://forum.hwbot.org/showpost.php?p=136877&postcount=37)

Whoah ... low 9 seconds!

Same like the two previous cups, older scores are allowed. Perhaps we can change that for the next Season. Discuss in Pro OC subforum?

Dat efficiency!

00 on the UD7 here meant CPU dead. But, I also had the situation where after testing a working CPU, that afterwards the 00 would be gone with the "dead" CPU. I think 00 might also mean CPU not recognised (microcode issue) or so.

Uploaded to Google Docs, this what you want? https://docs.google.com/file/d/0B2rJDSo8k6JYdWVQMXEtS2lqZ2c/

Well done Pixy! They did get the XTU 6c record too, so I guess that makes up for it. Nice scores from that event, by the way. Impressive stuff!

I am not aware of any page that would show that. I assume what you want is a list with the hardware (and global) rankings that generate the most points?

The HWBOT Hardware Index has nothing to do with the points. It's based on the average performance of hardware, so just the benchmarks.

Interesting ... Kingston HyperX 10th Anniversary 2x4GB DDR3-1866 CL9 1.65V Review Published by TaPaKaH There is actually a lot to be said about the influence of (competitive) overclocking on product development. Back when the processor operating frequency was still the main product differentiator, I bet a lot of companies felt cheated on by customers as well as industry members. Think back at that dual socket Pentium2 board from Abit that would allow you to run dual Celeron instead. The response was to lock as much down as possible, limiting the options for overclocking. Together with the lack of competition from AMD, Intel managed to raise the 'cost of overclocking' with their K/X-sku CPUs and Z/X-sku chipsets. In a way, that killed of the budget-limited enthusiasts. Which were actually the majority, I believe. People used to buy a lower SKU and then overclock it to the performance of the most high-end SKU. Competitive overclocking was certainly an active catalyst in this, as everyone shifted to the higher priced K/X-skus. Therefore enabling companies to go ahead with this approach. If that is what he's pointing to, I can agree to a certain extend.

Been watching part of the CEO Keynote at IDF, and saw this: No desktop, just mobile and all-in-one. :battle:

A couple of weeks back, we had a major database lockup because someone was using the "all benchmark" search functionality. If you have the server look through all 800,000 results, it takes a while. I don't think we can bring back the exact search functionality like before, but perhaps something where one search parameter is required and benchmark is one of the options.

It's not a bug, it is by design. RTC is used for systems that have no on-board clock generator. These systems are usually very low-end, where every added dollar matters. Windows 8 is designed to be compatible with all sorts of devices. I'm sure compatibility is higher on the priority list than at runtime overclocking.

Whoah, impressive stuff!

Welcome back!

Uh-oh!

Oh right, A88X is allowed since the only difference is official support for USB3, PCI-e Gen3, and Kaveri. Kaveri CPU is of course not allowed.