The Fastest Processor for Single Threaded Tasks

In the past we’ve had to make concessions for single-threaded application performance on modern day quad-core processors. For example, $266 will buy you two 3.33GHz cores or four 2.83GHz cores from Intel. I generally recommend going the quad-core option but there’s no getting around the fact that you do give up some performance when an application can’t take advantage of more than two threads.

With the Core i7 Extreme 975 the CPU can run at up to 3.60GHz when only one core is active (3.46GHz if more than one core is active). In my testing I found that the CPU almost always ran at its maximum turbo frequencies.

The graph below shows single-threaded performance in Cinebench R10. Note that while the Core 2 Duo E8600 (3.33GHz) was the top performer in this test for quite some time, the Core i7’s Turbo Mode has ensured that it’s no longer true.

Single Threaded Performance - Cinebench R10
The fastest single-threaded processors are now Intel's quad-core, eight-thread Core i7s

The biggest issue I see with the i7’s Turbo Mode today is that you only get one speed bin improvement (+133MHz) if 2 or more cores are active. The biggest boost (+266MHz) only comes when only a single core is active. Perhaps we’ll have to wait for Lynnfield for that.

The Test

Motherboard: Intel DX58SO (Intel X58)
Intel DX48BT2 (Intel X48)
MSI DKA790GX Platinum (AMD 790GX)
Gigabyte GA-MA790GP-DS4H (AMD 790GX)
Gigabyte GA-MA790FX-UD5P (AMD 790FX)
Chipset: Intel X48
Intel X58
Chipset Drivers: Intel (Intel)
AMD Catalyst 8.12
Hard Disk: Intel X25-M SSD (80GB)
Memory: G.Skill DDR2-800 2 x 2GB (4-4-4-12)
G.Skill DDR2-1066 2 x 2GB (5-5-5-15)
Qimonda DDR3-1066 4 x 1GB (7-7-7-20)
Corsair DDR3-1333 4 x 1GB (7-7-7-20)
Video Card: eVGA GeForce GTX 280
Video Drivers: NVIDIA ForceWare 180.43 (Vista64)
NVIDIA ForceWare 178.24 (Vista32)
Desktop Resolution: 1920 x 1200
OS: Windows Vista Ultimate 32-bit (for SYSMark)
Windows Vista Ultimate 64-bit
Index SYSMark 2007 Performance


View All Comments

  • strikeback03 - Wednesday, June 03, 2009 - link

    How many people bought a 940 or 965 anyway? Seems the vast majority of i7 sales were the 920, and then people OCed them. Guess we have to wait and see the OC results on the retail processors, but I'm guessing a lot more people would be happy to see a speed bumped 920 than these processors. Reply
  • aeternitas - Sunday, June 07, 2009 - link

    Take a look around at where you're, then ask yourself why you get that impression? Reply
  • rundll - Wednesday, June 03, 2009 - link

    Just wondering what happens to the Turbo Mode when you overclock a processor? Does it stop working, does it become obsolete? Is it useful only as long as you stick to stock specs?

    Anand said something about coming tests with retail i7-975. Is it possible to include some oc:ed i7-920 benchmarking results as well? That'd be great.
  • Gary Key - Wednesday, June 03, 2009 - link

    Turbo remains active during overclocking and depending on the BIOS, it remains active even if TDP limits are exceeded. I have the retail 975 coming and will provide overclocked numbers in a separate article centering on memory performance. Personally, buying a good 920 DO and overclocking it is your best option at this point unless you need a higher multiplier for extreme clocking, at which point the Xeon W3540 will provide you with the best bang for buck overclocks in this area. The unlocked multis on the 975 might allow a slightly better overclock or ability to clock memory, but maybe 0.02% of users will need it. I just have not found any general retail (not special binned from Intel) 975s that have out clocked the W3540s up high at this point and my retail 920 D0s just embarrass this 975 ES sample. Reply
  • rundll - Wednesday, June 03, 2009 - link

    Tks Gary, tks hemipowered.

    All clear now except that I just can't quite figure it out how the Turbo can pump up the volume after I've od:ed the cpu to max. Does it push it up to max+? Am I missing something here?

    If anyone sees it fit to answer, tks for that.
  • hemipowered - Wednesday, June 03, 2009 - link

    You can overclock and keep turbo on, I do Reply
  • hyvonen - Wednesday, June 03, 2009 - link

    I understand that gaming needs ATI/NVidia discrete cards, but just once I'd really like to see the power consumption numbers quoted without these power hogs...

    I'm interested in a quiet Media PC that plays BluRay/HD fine, does MP3/video encoding well, and could be used as a near-zero-load 24/7 file server (low idle power consumption). I have a PS3 for gaming; my PC would be fine with integrated graphics. But all the benchmarking setups are loaded with high-end graphics cards, massive PSUs and a ton of memory.

    None of the reviews show me how much the idle power of the CPU is, or that of the chipset/motherboard. I'm sure it's hard as hell to itemize the power consumptions of the components, but that's what I'm hoping Anand's team can figure out.

    I'd really like to see how AMD/ATI and Intel platforms work from performance/power point of view in low-power HTPC-like systems (with SSDs etc.). How much power does DDR2/DDR3 really consume? What about undervolting/clocking? Undervolting DDR3? Small-cache/small-size CPUs? Power/performance numbers are always for testbenches, never for CPUs themselves, and the leaky discrete graphics chips always mess up the results.

    I have a feeling that the Westmere dual-core desktop/mobile CPUs are going to be perfect for this sort of a system; 100W PicoPSU, SSD, WD Green HD... I just can't find the information anywhere.
  • strikeback03 - Wednesday, June 03, 2009 - link">

    too old to have i7 stuff in it, but you probably wouldn't want that in your system anyway.

    For reference, my desktop here at work has a 650W Enermax Infiniti PSU, Q6600@3GHz, P35 chipset, 1HDD, 1optical drive, 3 1GB sticks of DDR2, and a passive cooled nVidia 7300GT, and idles at just over 100W at the wall.

    To use a PicoPSU you probably need better info on what the components draw on each rail. When power is so limited it matters, I lost a M3-ATX due to that.
  • hyvonen - Wednesday, June 03, 2009 - link

    Thanks for the link - that article was great! I don't know how I missed it...

    I really hope they'll redo that stuff later with the new CPUs (especially the Westmere ones with IGP-in-the-package) and the new chipsets (P55 etc.)

    What happened with the PicoPSU/M3ATX?
  • strikeback03 - Thursday, June 04, 2009 - link

    It was never able to properly shut the system down, the jumper configuration was set to hibernate the system 5 seconds after the ignition was turned off, instead it waited about a minute. Once a week or so it would refuse to start the system when the ignition was turned on until I pulled the computer case out of the dashboard and then put it back in. Eventually that problem grew more frequent, now when power is applied the led on the M3 blinks but it won't turn a system on. I replaced it with an M2-ATX and that has not had any problems. I have an Intel Atom 330 LF2 board, that board seems to draw the majority of its power off the 5V rail, so apparently the 6A the M3 could provide was not enough 5V for long-term use. Reply

Log in

Don't have an account? Sign up now