Lucretius Posted October 18, 2010 Share Posted October 18, 2010 Are the values I see under Benchmark collected from users or are these just defaults that come as kind of a filler? If they're just filler, is there a way to update the filler with user submitted results? It would be great if there were a Validator type of repository similar to the one CPU-z has (see bottom), except for AIDA64 benchmarks/stress testing. If you could also integrate Linpack stress testing into the benchmarking this would be the only tool I would need. Such as: - Run the Linpack stress for 20 passes and add it to a system profile. - Run each of the benchmarks and add that to the system profile. - Submit system profile to central repository for comparison with other systems. ____________________________________ Quote Link to comment Share on other sites More sharing options...
Fiery Posted October 18, 2010 Share Posted October 18, 2010 The reference benchmark results listed on the benchmark pages are there for a reference. They are measured and collected by us (FinalWire), and they represent 100% reliable, valid, accurate results. We may come up with an online database with validator in the future, but our short-term plans focus on other improvements (automatic internet update, OpenCL-based GPGPU benchmarks + GPGPU stress test, etc). Thanks, Fiery Quote Link to comment Share on other sites More sharing options...
Lucretius Posted October 18, 2010 Author Share Posted October 18, 2010 All of those sound great. Thanks, Quote Link to comment Share on other sites More sharing options...
kzinti1 Posted October 27, 2011 Share Posted October 27, 2011 Every computer I've built, that I tested with Aida and Everest, has always been at the top of each benchmark. How is that possible? I'm no super computer builder or tweaker. I just slap together whatever parts I can afford and set the operating parameters to whatever works for me. How could I possibly build better computers than anyone else in the lists under the Benchmark tab? I am NOT that good! It seems as if the other computers ratings have been chosen to fall under my own ratings, thus not giving me an accurate representation of my own computers. I don't want some kind of an ego boost. I want actual results so I can start tweaking my computers to make them better than they actually are. IF I trusted just what Aida says, then I have no inducement whatsoever towards trying to better the operating environment of my systems. IF I choose to trust what the FutureMark programs indicate I should start selling off my computers piece by piece, use the money to buy myself a Dell and just forget about ever having a high-performance computer. The only Aida benchmarks I can actually trust are under the Tools tab, the Cache and Memory Benchmark and Aida64 CPUID. Both of which are confirmable using other programs. Under the Benchmark tab itself, all of the scores are not confirmable and, according to the various benchmark programs I own from FutureMark, do not have anything to do with the extraordinarily low scores I have with all of the FutureMark programs. Aida64 implies that my computers are absolutely top flight in performance, while the FutureMark benchmark programs say the complete opposite and cannot even compare to their lowest rated computers. Which am I to trust? Aida64 EE or FutureMarks 3DMark'03, '06, '11 and Vantage, plus, PCMark'05, PCMark Vantage 64 and PCMark 7? Aida64 EE says I have some of the best computers in existance, while the FutureMark programs say I have some of the worst. I'm not even going to mention the results from SiSoft's SANDRA program. I'm completely convinced that everybody there is quite insane. I'm almost there myself just trying to figure out if my computers are trash or treasure. Aida64 and FutureMark are either both sides of schizophrenia or a case of manic/depression. Quote Link to comment Share on other sites More sharing options...
Fiery Posted October 28, 2011 Share Posted October 28, 2011 I'm not sure what's wrong with the computer you benchmark, but many hardware reviewers from review portals think AIDA64 as a diagnostic software, and the benchmarks it offers are very useful and properly represent CPU and memory performance: http://www.aida64.com/testimonials Maybe if you could show us actual screen shots of benchmark results obtained with AIDA64 would help. Make sure to also explain what system it is, and whether it is overclocked or not. Please note that the reference benchmark results are always obtained on non-overclocked and non-tweaked systems, to represent "out of the box" performance of average PC systems. Regards, Fiery Quote Link to comment Share on other sites More sharing options...
Med_spravka Posted October 29, 2012 Share Posted October 29, 2012 Hi I am from Russia. Why the model i5 460m results: CPU QUEEN - 17696, and the model i5 2450m results: CPU QUEEN - 9899? Quote Link to comment Share on other sites More sharing options...
Fiery Posted October 29, 2012 Share Posted October 29, 2012 Hi I am from Russia. Why the model i5 460m results: CPU QUEEN - 17696, and the model i5 2450m results: CPU QUEEN - 9899? It's not easy to post judgements on that anomaly without knowing a bit more details of the i5-2450M system. Our best guess would be that either a core, or HyperThreading is disabled on that system, or maybe power management settings block your processor to work at full performance. Quote Link to comment Share on other sites More sharing options...
Med_spravka Posted October 29, 2012 Share Posted October 29, 2012 It's not easy to post judgements on that anomaly without knowing a bit more details of the i5-2450M system. Our best guess would be that either a core, or HyperThreading is disabled on that system, or maybe power management settings block your processor to work at full performance. I expect from a turbo boost. At i5 460m - turbo boost 1.0 and a i5 2450m - turbo boost 2.0. Maybe? Quote Link to comment Share on other sites More sharing options...
Fiery Posted October 29, 2012 Share Posted October 29, 2012 Turbo Boost would help to make scores higher, but even without it, i5-2450M should perform at least equal to a i5-460M. Please verify that all cores are enabled on the i5-2450M system, and HyperThreading is also enabled. Please also right-click on the bottom status bar of AIDA64 main window --> CPU Debug --> CPUID & MSR Dump. Copy-paste the full results into this topic. Quote Link to comment Share on other sites More sharing options...
Med_spravka Posted October 30, 2012 Share Posted October 30, 2012 Turbo Boost would help to make scores higher, but even without it, i5-2450M should perform at least equal to a i5-460M. Please verify that all cores are enabled on the i5-2450M system, and HyperThreading is also enabled. Please also right-click on the bottom status bar of AIDA64 main window --> CPU Debug --> CPUID & MSR Dump. Copy-paste the full results into this topic. For example, model i7 2670QM , the results of the CPU Queen - 14371. Quote Link to comment Share on other sites More sharing options...
Med_spravka Posted October 31, 2012 Share Posted October 31, 2012 Sorry, I already understood. There were different versions Quote Link to comment Share on other sites More sharing options...
Med_spravka Posted November 5, 2012 Share Posted November 5, 2012 Why the model i5 2450m (4gb memory 1333 Mhz) results: Memory read - 16323 Mb/s, and the model i7 3610QM (4gb memory 1600 Mhz) results: Memory read - 11989 Mb/s ? Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.