Just for curiosity... I've OC'ed i7 860 to 4 Ghz. Results: Looking at benchmarks results "agreggate", is it normal that Core i5 -750@2'6 Ghz scored same as i7 860@2'8 Ghz ? And when it's OC'ed to 4.33Ghz it scored much higher? PS: Judging from a few overclock-Corei5 guides I'd say 4.33Ghz is quite extreme....
Good point. The Core i5-750 (2.66 GHz) was also faster than the Core i7-920 (2.66 GHz) in Pass 1. But it was quite different in Pass 2. The Core i5-750 was much slower. Look at the Average / Aggregate score. The Core i5-750 ended up slower than the Core i7-920 and the Core i7-860. Although cheating is possible, we hope that no one does this as we merely want to provide a record of performance results for various CPUs at different clock speeds.
x264 HD BENCHMARK 3.0 RESULTS Please do not compare it with older versions of the benchmark! Please copy/paste everything below the line to to report your data ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Results for x264.exe r1342 -------------------------- encoded 1442 frames, 110.16 fps, 3901.21 kb/s encoded 1442 frames, 111.76 fps, 3900.68 kb/s encoded 1442 frames, 112.11 fps, 3900.68 kb/s encoded 1442 frames, 109.93 fps, 3900.68 kb/s encoded 1442 frames, 33.17 fps, 3971.56 kb/s encoded 1442 frames, 33.24 fps, 3970.61 kb/s encoded 1442 frames, 33.36 fps, 3970.65 kb/s encoded 1442 frames, 33.25 fps, 3968.26 kb/s System Details -------------- Name Intel Core i7 920 product System Product Name Codename Bloomfield Specification Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz Specification PC3-10700 Specification PC3-10700 Specification PC3-10700 Specification PC3-10700 Specification PC3-10700 Specification PC3-10700 Core Stepping D0 Technology 45 nm Stock frequency 2666 MHz Core Speed 4438.4 MHz Northbridge Intel X58 rev. 13 Southbridge Intel 82801JR (ICH10R) rev. 00 CAS# latency (CL) 9.0 RAS# to CAS# delay (tRCD) 9 RAS# Precharge (tRP) 9 Cycle Time (tRAS) 24 Command Rate (CR) 2T Command Rate 2T Command Rate 2T Command Rate 2T Command Rate 2T Command Rate 2T Command Rate 2T Memory Frequency 845.4 MHz (2:8) Memory Type DDR3 Memory Size 12288 MBytes Channels Triple Windows Version Microsoft Windows 7 (6.1) Ultimate Edition (Build 7600) Number of processors 1 Number of threads 4 Number of threads 4 (max 16) L2 cache 4 x 256 KBytes, 8-way set associative, 64-byte line size Instructions sets MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, EM64T Package (platform ID) Socket 1366 LGA (0x1) Temperature 0 33°C (91°F) [0x21] (SYSTIN) Temperature 1 55°C (130°F) [0x6D] (CPUTIN) Temperature 2 13°C (54°F) [0x19] (AUXTIN) Temperature 0 68°C (154°F) [0x20] (Core #0) Temperature 1 65°C (148°F) [0x23] (Core #1) Temperature 2 61°C (141°F) [0x27] (Core #2) Temperature 3 64°C (147°F) [0x24] (Core #3)
A quick note to the above run. I run my i7 920 overclocked without Hyperthreading, as it runs much cooler, requires less voltage for a stable overclock (1.4v on the 920 Stepping D0 above for 4.43ghz), and from what I can tell HT is a performance advantage in very few applications. The motherboard is an Asus Rampage II GENE, BIOS 0904 (stay away from 1033 and 1104), BCLK 211 @ 21x. The memory is Patriot PVT312G1333LLHK @ 1690mhz VDIMM @ 1.65v. The CPU cooler is a Coolermaster V10, in a Coolermaster Storm Sniper Black case (plenty of cooling). I was looking for tools for testing HD video encoding, and ran across your benchmark. Thanks for posting it! I will check back and see how the testing is going, and check if anyone has any questions.
I noticed that the download link provided by the benchmark application pointed to a download for the Avisynth_258.exe installer, while the link in the article points to the Avisynth_257.exe installer. My run was done with the link given by the benchmark, Avisynth_258.exe Scott
Here is an i7 920 @ 2.67ghz at stock settings throughout. Turbo disabled, and Hyperthreading enabled. x264 HD BENCHMARK 3.0 RESULTS Please do not compare it with older versions of the benchmark! Please copy/paste everything below the line to to report your data ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Results for x264.exe r1342 -------------------------- encoded 1442 frames, 63.98 fps, 3899.02 kb/s encoded 1442 frames, 63.46 fps, 3899.02 kb/s encoded 1442 frames, 64.60 fps, 3898.50 kb/s encoded 1442 frames, 64.01 fps, 3899.02 kb/s encoded 1442 frames, 24.81 fps, 3971.06 kb/s encoded 1442 frames, 25.01 fps, 3954.02 kb/s encoded 1442 frames, 25.10 fps, 3959.75 kb/s encoded 1442 frames, 24.85 fps, 3971.04 kb/s System Details -------------- Name Intel Core i7 920 product System Product Name Codename Bloomfield Specification Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz Specification PC3-10700 Specification PC3-10700 Specification PC3-10700 Specification PC3-10700 Specification PC3-10700 Specification PC3-10700 Core Stepping D0 Technology 45 nm Stock frequency 2666 MHz Core Speed 2669.8 MHz Northbridge Intel X58 rev. 13 Southbridge Intel 82801JR (ICH10R) rev. 00 CAS# latency (CL) 9.0 RAS# to CAS# delay (tRCD) 9 RAS# Precharge (tRP) 9 Cycle Time (tRAS) 24 Command Rate (CR) 2T Command Rate 2T Command Rate 2T Command Rate 2T Command Rate 2T Command Rate 2T Command Rate 2T Memory Frequency 801.0 MHz (2:12) Memory Type DDR3 Memory Size 12288 MBytes Channels Triple Windows Version Microsoft Windows 7 (6.1) Ultimate Edition (Build 7600) Number of processors 1 Number of threads 8 Number of threads 8 (max 16) L2 cache 4 x 256 KBytes, 8-way set associative, 64-byte line size Instructions sets MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, EM64T Package (platform ID) Socket 1366 LGA (0x1) Temperature 0 30°C (85°F) [0x1E] (SYSTIN) Temperature 1 39°C (102°F) [0x4E] (CPUTIN) Temperature 2 16°C (59°F) [0x1F] (AUXTIN) Temperature 0 48°C (118°F) [0x34] (Core #0) Temperature 1 44°C (111°F) [0x38] (Core #1) Temperature 2 47°C (116°F) [0x35] (Core #2) Temperature 3 42°C (107°F) [0x3A] (Core #3)
Here is the same i7 920 @ 2.67ghz at stock settings, without Hyperthreading )HT) enabled. Note differences in results for Pass 1 (faster without HT), and Pass 2 (faster with HT). x264 HD BENCHMARK 3.0 RESULTS Please do not compare it with older versions of the benchmark! Please copy/paste everything below the line to to report your data ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Results for x264.exe r1342 -------------------------- encoded 1442 frames, 70.60 fps, 3900.68 kb/s encoded 1442 frames, 70.52 fps, 3900.68 kb/s encoded 1442 frames, 70.29 fps, 3900.68 kb/s encoded 1442 frames, 70.07 fps, 3900.68 kb/s encoded 1442 frames, 20.61 fps, 3971.32 kb/s encoded 1442 frames, 20.45 fps, 3972.12 kb/s encoded 1442 frames, 20.63 fps, 3971.77 kb/s encoded 1442 frames, 20.51 fps, 3970.80 kb/s System Details -------------- Name Intel Core i7 920 product System Product Name Codename Bloomfield Specification Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz Specification PC3-10700 Specification PC3-10700 Specification PC3-10700 Specification PC3-10700 Specification PC3-10700 Specification PC3-10700 Core Stepping D0 Technology 45 nm Stock frequency 2666 MHz Core Speed 2669.9 MHz Northbridge Intel X58 rev. 13 Southbridge Intel 82801JR (ICH10R) rev. 00 CAS# latency (CL) 9.0 RAS# to CAS# delay (tRCD) 9 RAS# Precharge (tRP) 9 Cycle Time (tRAS) 24 Command Rate (CR) 2T Command Rate 2T Command Rate 2T Command Rate 2T Command Rate 2T Command Rate 2T Command Rate 2T Memory Frequency 801.0 MHz (2:12) Memory Type DDR3 Memory Size 12288 MBytes Channels Triple Windows Version Microsoft Windows 7 (6.1) Ultimate Edition (Build 7600) Number of processors 1 Number of threads 4 Number of threads 4 (max 16) L2 cache 4 x 256 KBytes, 8-way set associative, 64-byte line size Instructions sets MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, EM64T Package (platform ID) Socket 1366 LGA (0x1) Temperature 0 31°C (87°F) [0x1F] (SYSTIN) Temperature 1 38°C (100°F) [0x4C] (CPUTIN) Temperature 2 15°C (58°F) [0x1D] (AUXTIN) Temperature 0 47°C (116°F) [0x35] (Core #0) Temperature 1 45°C (112°F) [0x37] (Core #1) Temperature 2 46°C (114°F) [0x36] (Core #2) Temperature 3 44°C (111°F) [0x38] (Core #3)
Thanks for the results, hmscott! +rep! It is certainly interesting to note the effect of turning off HT on the results. Regarding AVISynth, the newer version has no effect on the results.
x264 HD Benchmark 3.08 The new x264 HD Benchmark 3.0 is here! This is the improved version of the x264 HD Benchmark 2.0 benchmark which has become a standard CPU benchmark for Tech ARP as well as many other hardware enthusiasts. The new x264 HD Benchmark 3.0 ups the ante by using the latest r1342 x264 encoder as well as other changes under the hood. This makes the x264 HD Benchmark 3.0 the new best tool for testing the full capabilities of modern multi-core processors. Do note that due to the changes, the results from x264 HD Benchmark 3.0 are NOT comparable with results from older versions of the benchmark. In addition to the release of the benchmark, we will be compiling benchmark results for this new version. So please contribute your new benchmark results to this project! Check out the latest results! Link : x264 HD Benchmark 3.08
Thanks for the +rep Adrian! And, yes I am also now more interested in trying other applications with and without HT enabled to see what the effect is on the performance. I wonder if there are optimizations for both that can be included in applications to take advantage of either situation. Scott
x264 HD BENCHMARK 3.0 RESULTS Please do not compare it with older versions of the benchmark! Please copy/paste everything below the line to to report your data ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Results for x264.exe r1342 -------------------------- encoded 1442 frames, 101.69 fps, 3899.02 kb/s encoded 1442 frames, 101.71 fps, 3899.02 kb/s encoded 1442 frames, 101.25 fps, 3899.02 kb/s encoded 1442 frames, 101.39 fps, 3898.30 kb/s encoded 1442 frames, 39.73 fps, 3949.41 kb/s encoded 1442 frames, 40.10 fps, 3959.88 kb/s encoded 1442 frames, 39.83 fps, 3966.32 kb/s encoded 1442 frames, 39.70 fps, 3968.95 kb/s System Details -------------- Name Intel Core i7 920 product System Product Name Codename Bloomfield Specification Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz Specification PC3-10700 Specification PC3-10700 Specification PC3-10700 Specification PC3-10700 Specification PC3-10700 Specification PC3-10700 Core Stepping D0 Technology 45 nm Stock frequency 2666 MHz Core Speed 4334.1 MHz Northbridge Intel X58 rev. 13 Southbridge Intel 82801JR (ICH10R) rev. 00 CAS# latency (CL) 9.0 RAS# to CAS# delay (tRCD) 9 RAS# Precharge (tRP) 9 Cycle Time (tRAS) 24 Command Rate (CR) 2T Command Rate 2T Command Rate 2T Command Rate 2T Command Rate 2T Command Rate 2T Command Rate 2T Memory Frequency 825.5 MHz (2:8) Memory Type DDR3 Memory Size 12288 MBytes Channels Triple Windows Version Microsoft Windows 7 (6.1) Ultimate Edition (Build 7600) Number of processors 1 Number of threads 8 Number of threads 8 (max 16) L2 cache 4 x 256 KBytes, 8-way set associative, 64-byte line size Instructions sets MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, EM64T Package (platform ID) Socket 1366 LGA (0x1) Temperature 0 30°C (85°F) [0x1E] (SYSTIN) Temperature 1 61°C (140°F) [0x79] (CPUTIN) Temperature 2 17°C (61°F) [0x21] (AUXTIN) Temperature 0 72°C (161°F) [0x1C] (Core #0) Temperature 1 72°C (161°F) [0x1C] (Core #1) Temperature 2 71°C (159°F) [0x1D] (Core #2) Temperature 3 64°C (147°F) [0x24] (Core #3)
BTW, hmscott, can you redo your benchmark @ 4.43 GHz (posted here - http://forums.techarp.com/reviews-articles/25637-x264-hd-benchmark-3-0-a-9.html#post355999) but this time with HT enabled? BTW, are these results with HT enabled and Turbo Boost disabled? For directly comparable results, I think we will only publish results for Core i7/i5/i3 processors with HT enabled and Turbo Boost disabled. Thanks!
Why is my HTPC so slow? Results for x264.exe r1342 -------------------------- encoded 1442 frames, 28.43 fps, 3899.26 kb/s encoded 1442 frames, 28.48 fps, 3899.26 kb/s encoded 1442 frames, 28.33 fps, 3899.26 kb/s encoded 1442 frames, 26.76 fps, 3899.26 kb/s encoded 1442 frames, 6.83 fps, 3971.79 kb/s encoded 1442 frames, 6.84 fps, 3971.71 kb/s encoded 1442 frames, 6.85 fps, 3971.56 kb/s encoded 1442 frames, 6.84 fps, 3971.51 kb/s System Details -------------- Name Intel Core 2 Duo E8400 Codename Wolfdale Specification Intel(R) Core(TM)2 Duo CPU E8400 @ 3.00GHz Core Stepping C0 Technology 45 nm Stock frequency 3000 MHz Core Speed 1999.7 MHz FID range 6.0x - 9.0x Northbridge Intel X38 rev. 01 Southbridge Intel 82801IR (ICH9R) rev. 02 CAS# latency (CL) 5.0 RAS# to CAS# delay (tRCD) 5 RAS# Precharge (tRP) 5 Cycle Time (tRAS) 15 Command Rate (CR) 2T Memory Frequency 333.3 MHz (1:1) Memory Type DDR2 Memory Size 2048 MBytes Channels Dual, (Symmetric) Windows Version Microsoft Windows 7 (6.1) Ultimate Edition (Build 7600) Number of processors 1 Number of threads 2 Number of threads 2 (max 2) L2 cache 6144 KBytes, 24-way set associative, 64-byte line size Instructions sets MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, EM64T Package (platform ID) Socket 775 LGA (0x0) Temperature 0 n.a. [0x20] (TMPIN0) Temperature 1 n.a. [0x5E] (TMPIN1) Temperature 0 42°C (107°F) [0x2A] (CPU) Temperature 1 28°C (82°F) [0x1C] (System) Temperature 2 43°C (109°F) [0x2B] (PWM Phase 1) Temperature 3 43°C (109°F) [0x2B] (PWM Phase 2) Temperature 4 44°C (111°F) [0x2C] (PWM Phase 3) Temperature 5 43°C (109°F) [0x2B] (PWM Phase 4) Temperature 6 41°C (105°F) [0x29] (PWM Phase 5) Temperature 0 46°C (114°F) [0x36] (Core #0) Temperature 1 42°C (107°F) [0x3A] (Core #1)
Adrian, at 4.43ghz with HT on the temps go too high, I am not using water cooling, so it looks like for now 4.334ghz is the best speed possible with HT on with air cooling. Even at 4.334ghz, I wouldn't want to run at 100% for long, even though it is stable. HT at high overclock speeds really adds on the heat, 10-20c+ depending on the application running. For longer cpu life, in my configuration, I wouldn't run HT on 100% at higher than 4.2ghz for long application runs (24/7). For this benchmark, where HT on/off adds performance to one pass, but takes it away in another, it would be nice to see if the overall elapsed time increases, decreases, or if it is essentially a wash. Do you think the benchmark could be modified to give an overall elapsed time for both passes and the entire benchmark run? iXBT labs had a series of articles benchmarking various elements for the i7, and Part 3 covered Turboboost and Hyperthreading. How CPU Features Affect CPU Performance, Part 3 iXBT Labs - How CPU Features Affect CPU Performance, Part 3 - Page 1: Introduction, tests, part 1 It may be that for a particular application mix, moving from 4.43ghz with HT off to 4.0ghz with HT on might give noticeably better performance. For me, running at the higher clock with HT off, gives a noticeable performance improvement. And, at the same clock rate, except for benchmarks, I don't notice an improvement with HT on. BTW, I can't find the other posting for 4.33ghz, I wanted to see what their temps were, and other settings. Can you post it in the thread? Thanks Adrian!, Scott
The 4.43ghz run was done with HT OFF, and there is no Turboboost involved when overclocking at the max multiplier of 21x for the i7 920. In the Asus Rampage II BIOS (0904) there is no "Turbo-Tech" setting while in "Manual" OC mode. I run the multi at 21x, and vary the speed via the BCLK setting (from 133 stock to 216), and therefore I don't see Turboboost effects. In that run I was already at 21x @ BCLK 211. For lower speeds runs, I use i7Turbo.exe from Real Temp to "Turbo Mode Disable" from Windows to turn off Turboboost. techPowerUp! :: Download Real Temp 3.40 When running with HT ON at 4.43ghz, the CPU would down clock during Pass 2, changing the multiplier from 21x to 20x, dropping speed to 4.220ghz due to overheating. I don't consider that a clean run This is the BIOS setting that protects the CPU by throttling speed: CPU TM Function - Enabled <= Overtemp Protection for CPU That is one setting you do not want to turn off! Unless you are going for a record, and have a box of spare CPU's on hand The C1E/Intel SpeedStep/C-State BIOS settings are Enabled. During a benchmark run you can disable their effects by changing the "Power Options" control panel "Advanced Power" => "Processor power management" => "Minimum processor state Setting" to 100% when benchmarking. I set it to 5% for power savings, when I am running normally. Some configurations have trouble with C1E/Intel SpeedStep/C-State settings Enabled when overclocked, and turning them off is one of the first things you do for stability when testing for MAX BCLK and MAX multiplier. The next step would be to move to water cooling to stop the CPU throttling at 4.43ghz with HT ON, but for now I am happy at 4.43ghz with HT OFF, and 4.33ghz with HT on.
Some processor problem My BIOS settings are at 9x and 3000MHz, but my Everest diagnostic showed me this same lack as the x-test... My HTPC seems to be running with 6x and 1999,7MHz. How is this possible? Can anyone help find the lack in processor speed? Everest diagnostic doesn't report anything to be out of order here. The voltage readings for processor are: CPU VID 1,1125V CPU CORE 1,11V CPU VTT 1,11V I already tested with some light overclocking too, that didn't help. BIOS settings were at 9x and 3120MHz, Everest showed me 6x and 2040MHz, so it seems to lack exactly 1/3 of the clock. (6x vs 9x).