About HWBOT Testlab – HWBOT Content Site

About HWBOT Testlab

1. Philosophy behind HWBOT Testlab

Nowadays overclocking is a very interesting tool to demo a new product. Marketers enjoy using “the fastest” and “world record” claims because it’s an easy way to attract attention to their new board as well as point out exactly how much better they are than the competition. Even though a (small) group within this community has an intrinsic hate towards anything that comes from any vendor, I wouldn’t say having the industry involved in overclocking is a bad thing per se. If the industry can be engaged to support and feed the overclocking community with not only products designed for overclocking (PWM, OC buttons, …) but also fix overclocking related issues (BIOS!), we can squeeze more out of our system. The marketing and advertisement is just part of the cycle: they feed us, we feed them.

Marketing claims become a problem when there is no information that validates the claim. Without going into too much detail on who and when, I must say that over the past years I’ve seen several media ‘confirming’ overclocking records without having any proof. No validation link, no screenshot. Perhaps even worse than that is the severe lack of information regarding these overclocking claims. How was the result obtained? What BIOS? What settings? How can I reproduce this result at home? Many questions, but very few answers.

In a similar way as companies like to use overclocking records for marketing purposes, they advertise general performance or overclocking capabilities on the product boxes or marketing material. Just look at some of the examples below.

In no way is it our intention to put a question mark to any of these performance or overclocking claims. Although there might be some exception, in general we believe that the internal testing and the results that come from that testing are per definition invalid. But, the lack of detailed test configuration settings (eg: voltage, frequency, temperature) makes it difficult to reproduce and validate that claim. Aiming for transparency, it would be a lot better to have full testing reports done by a 3rd party lab.

The HWBOT Testlab has been established for exactly that: test and verify internal overclocking-related results.

2. Rules of Engagement

Having a 3rd party validating overclocking-related performance claims is only useful if the 3rd party actually improves the transparency. If the 3rd party merely sticks to saying ‘yes’ or ‘no’, it’s as useful and trustworthy as the results coming from the internal testing. So, in order to improve the transparency we have our Rules of Engagement. The RoE is a set of principles that we stick to for any HWBOT Testlab result. Note that the RoE are not a definitively set in stone and may change over time.

  • #1: HWBOT Testlab Does Not Guarantee Performance or Overclocking Results.
    Even though it would be highly preferable for vendors to just ring up the Testlab and get any marketing claim verified, the Testlab does not commit to anything but testing the platform. If the end result is not satisfactory, so be it.
  • #2: Testlab Is Not WR Machine
    The HWBOT Testlab exists only to provide transparency and additional validity to overclocking and performance claims from the industry. It is not our task, nor our mission to provide any vendor with a record. The HWBOT Testlab is not a competitive overclocking team; it merely serves the purpose of validating a specific overclocking result.
  • #3: Full Disclosure of Test Methodology and Tools Used.
    At all times, any HWBOT Testlab Verified Result will have full documentation on how the result was obtained. In addition, all the tools that were used to obtain the result will be made available to the public. This includes, for example, the used BIOS version, any special software for voltage or frequency control and full disclosure on the used BIOS settings. If the vendor does not agree to full disclosure, the obtained result will not be HWBOT Testlab Verified.
  • #4: The Testlab Conducts Own Testing.
    The HWBOT Testlab consists only of official HWBOT employees and allows no outside interference in terms of obtaining a specific result. Outside help (eg: BIOS engineer) may be requested by the Testlab, in which case interference is allowed. Do note that any outside interference will also be documented in the Testlab report.
  • #5: HWBOT Testlab Verified Usage.
    In exchange for the full documentation, the vendor may use the HWBOT Testlab Verified logo. Using the logo is not mandatory. Every logo will have its own Testlab Report page (see below).
  • 3. HWBOT Testlab Verified Results

    Testlab IDVendorHardwareBenchmarkTestlab Verified ResultSubmission LinkTestlab Report
    #1GIGABYTEF2A85X-UP43DMark Vantage10000+linklink

    4. More information / Questions

    If you have any questions or concerns regarding any other HWBOT Testlab report, or the HWBOT Testlab concept in general, feel free to get in touch through email or the forums. We are always interested in hearing constructive criticism and new ideas to improve the validity of the Testlab results.

Testlab articles:

SkatterBencher #3: Overclocking the ROG GX700, a Water-Cooled Notebook

  • Guides, News, Articles, testlab
  • 5

Hello SkatterBenchers. Today we will show you how to overclock this notebook in a minimum amount of steps and time. This is the ASUS ROG GX700 V0 notebook which comes with watercooling as well as a Core i7 6820HK processor, a desktop grade GTX 980 graphics card and 64GB of memory. To do the overclocking we will be using the ROG Gaming Center application and in there we will switch up the Turbo Gear settings. We will be using four different benchmarks in the operating system to measure the performance; Intel XTU, ROG Realbench, F1 2014 and 3DMark Fire Strike Extreme. Note: Today will be a little bit different than our other SkatterBencher videos because we can’t really change that much manually.

As a final step of our overclocking adventure, we will switch to manual mode. We increased the CPU frequency to 4.0 GHz, we increased the CPU cache frequency to 4GHz. We enabled XMP and then we also increased the GPU core frequency by 200MHz as well as the GPU memory frequency by 300MHz as well. In a final round of benchmarking we find that the XTU performance has gone up to 1,271 points which is 18% over stock performance. In Fire Strike Extreme we see that our performance went up to 6,665 points which almost 40% over stock performance - this is thanks to the 1,427MHz GPU frequency and 3.8GHz memory frequency. In F1 2014 our minimum FPS has gone up to 71, which is 20% over stock performance and in ROG Realbench we now have a score 103,741 points, 23% over stock. The CPU temperature under load is 83C and the GPU temperature under load is only 41C.

Overclocking Like In The Olden Days: Easy 50% Free Performance Increase with AMD Kaveri

  • News, Editorials, testlab
  • 0

50% extra performance. Just on air cooling - no LN2 required. Yes, ambient cooling is all it takes to significantly increase the performance of AMD's new APU codenamed Kaveri. To make things short, what makes the Kaveri so interesting is the HSA technology. HSA allows both CPU and GPU to read from the same memory registry, without having to reserve it for one specific purpose. In previous APU generations we already noticed how important system memory overclocking is for the 3D performance of the APU. Although AMD did not really highlight overclocking as one of the major selling points of the Kaveri APU, we are convinced the performance increase gained through overclocking is very significant.

It brings us back to the early days of overclocking, where a fiddling in the BIOS resulted in great and free performance. Continue reading ->