Intel Core i3-2100 LGA 1155: Sandy Bridge dual core! - BeHardware
Written by Guillaume Louel and Marc Prieur
Published on January 31, 2011
At CES in Las Vegas, Intel announced a vast range of 32nm LGA 1155 processors, including no fewer than fourteen desktop and fifteen mobile processors. However, although the four core versions have been officially available since January 9th, the dual cores only started arriving in stores on February 20th. We have, of course, managed to test them!
Sandy Bridge architectureSandy Bridge is Intel’s code name for its new architecture and the processors based on it. This new generation is a ‘tock’ in Intel parlance, signifying a new architecture using a manufacturing process already used previously. Sandy Bridge CPUs are therefore manufactured at 32nm, like the Westmeres, the 32nm versions of Nehalem.
Among the innovations introduced by Sandy Bridge CPUs, the most notable are:
- A new Socket LGA 1155
- An integrated IGP sharing the L3 cache, now known as LLC (Last Level Cache)
- An improvement of the IPC and performance per watt
- A new AVX (Advanced Vector Extension) instruction set
- A new version of Turbo Boost
Beyond these aspects of the spec on which Sandy Bridge differs from its predecessor, the technological choices introduced or reintroduced with Nehalem have been retained, namely:
- The integrated DDR3 memory controller (dual channel and DDR3-1333)
- The built-in PCI-Express controller (2 x 8 PCI-E 2.0)
- A three levels cache architecture
We won’t go back over the architecture here but for more detail you can consult our report on the LGA 1155 socket Core i7s and i5s.
Sandy Bridge versus Clarkdale, the range
Sandy Bridge versus ClarkdaleThe dual core version of Sandy Bridge for desktop is equipped with an HD 2000 type IGP. The die is 131mm˛ and has 504 million transistors, while the mobile version has an HD 3000 IGP and measures 149mm˛ with 624 million transistors.
For comparison, the previous dual core LGA 1156 Core i3/i5s (the Clarkdales) combine two dies, the first being the CPU engraved at 32nm on 81mm˛ with 383 million transistors and the second the northbridge with an IGP with a 114mm˛ die and 177 million transistors engraved at 45nm.
Clarkdale, then, was a departure from the monolithic Intel designs. By grouping everything on the same 32nm die, Intel has reduced manufacturing costs further, with total die area down from 195mm˛ to 149mm˛. Abandoning the 45nm manufacturing for 32nm also reduces energy consumption. You’ll have noticed that the number of transistors is also down overall, dropping from 560 to 504 million. This is in part due to the smaller L3 / LLC cache, which is down from 4 MB on the Clarkdales to 3 MB on the dual core Sandy Bridges.
Putting the memory controller on the same die as the CPU is also a big advantage in terms of performance. Putting it on the northbridge, as was the case with Clarkdale, meant it was rather slow, with DDR3-1333 latency up by 64% (75ns against 45ns) and bandwidth down by 37% (15.5 GB/s against 9.5 GB/s) in comparison to Lynnfield (quad core 45nm LGA 1156). This problem has therefore been resolved on the LGA 1156 dual core Sandy Bridge CPUs.
The Core i3 (and i5!) dual core range
The Sandy Bridge dual core range is pretty small, made up of just two standard models:
- Core i3-2100, 3.1 GHz, 3 MB LLC, HD Graphics 2000, 65W, $117
- Core i3-2120, 3.3 GHz, 3 MB LLC, HD Graphics 2000, 65W, $138
For comparison AMD’s Athlon II X4 645, 3.1 GHz, is listed at $112 and its Phenom II X4 955 at $145. In the Intel LGA 1156 range, the i3-560 (3.33 GHz) is listed at $138, against $117 for the i3-550 (3.2 GHz).
Another model with reduced TDP is however exclusively designed for PC manufacturers:
- Core i3-2100T, 2.5 GHz, 3 MB LLC, HD Graphics 2000, 35W
Lastly, there is also one dual core Core i5! The difference with the i3s is that you also get Turbo mode. The LGA 1155 Core i5 range, like LGA 1156, includes CPUs with 2 physical / 4 logical cores as well as 4 physical cores, the common denominator being the Turbo. Naturally, we feel this is pretty misleading!
- Core i5-2390T, 2.7 GHz, Turbo 3.5 GHz, 3 MB of LLC, HD Graphics 2000, 35W
In contrast to the Core i3s, this Core i5-2390T also supports Intel TXT, Intel VT-d (VT-x is supported on i3) and AES-NI instructions. Given that all these processors use a common die, the limitations on the Core i3s are purely for commercial reasons and are therefore questionable.
The Core i3 2100 (energy consumption, o/c)
The Core i3-2100For this test, we used a Core i3-2100, the cheapest version, clocked at 3.1 GHz:
Note that the power supply voltage is 1.15V, against 1.2V for the standard Core i5 2300/2400/2500s or 1.1V for the "S" versions.
Energy consumptionWe measured the power consumption of the configuration at the wall socket of the power supply used, with a yield of around 80%. For the test in load we used Prime95. This means that other components such as the graphics card or the hard drive are in idle when these readings are taken.
Hold the mouse over the graph to see a classification of the CPU’s by result.
In load the Core i3-2100 configuration is clearly more economical, drawing 7 watts less than with the Core i3-540.
Here now is the reading for power consumption at the ATX12V, using a clip ammeter. We only give the results for LGA 1155 CPUs here as no exact comparison is possible from one platform to another.
At under 35 watts, the power draw at the ATX12V is impressive. We haven’t seen such a low power draw since the 45nm Socket 775 dual cores, with the major difference that those CPUs didn’t have the memory controller on the same die!
OverclockingAs discussed in our Core i5 / i7 LGA 1155 report, Intel now controls the overclocking capabilities on this platform. Overclocking via the bus is very limited and you’re therefore limited to the multiplier.
Unfortunately, the multiplier is only unlocked on the ‘K’ version processors. On non-K CPUs with Turbo, you can push the Turbo multipliers to the maximum and increase power consumption limits to clock your processor up by 500 MHz on the base clock with 4 cores used.
The bad news for the Core i3s is that they don’t exist as ‘K’ versions and don’t have Turbo. Nothing can therefore be done with the multipliers.
All that’s left is the bus clock, which is at 100 MHz by default. On the Core i3-2100 we couldn’t get any higher than 104 MHz stable, or 103.8 MHz to be exact. This gave us a clock of 3.216 GHz up from 3.1 GHz … wow!
The platformsFor this test, we used our usual protocol, as detailed at the bottom of this page in the report on LGA 1155 Core i5 and i7 .
Intel HD Graphics 2000
Intel HD Graphics 2000Like the other Sandy Bridge processors, the Core i3 2100 has an integrated graphics core. This is the HD 2000 with six execution units. Note that although the Core i3 2100 doesn’t have Turbo mode for its CPU cores, the graphics core does have Turbo mode which allows it to clock up to 1100 MHz (against a base clock of 850 MHz) which is identical on paper to the graphics core in the Core i5 2500. One of the particularities of Sandy Bridge architecture is that the last level cache is shared with the graphics core. Given the fact that it has been reduced in size here (3 MB) we wanted to check the impact on performance. For more details on the Sandy Bridge HD Graphics architecture, we refer you to our previous article.
Energy consumptionWe measure the energy consumption of the graphics core using the internal reading on the Sandy Bridge Power Control Unit via the hwinfo32 software:
No surprises here, at 8.5 watts in load, the HD 2000 is particularly economical. Note that we’re a long way from the TDP announced for the combined graphics and CPU load.
3D performanceWe measured the 3D performance of the Core i3 2100. Although the IGP is identical in its configuration to the solution on the Core i5 2500, we wanted to measure the impact, if there is one, of the reduced size LLC. We also tested the Core i3 2100 together with the Radeon HD 5450 (GDDR3).
Here are the platforms used, with Windows 7 64 bits being the OS used across all platforms:
- Intel Core i5 661, Asus P7H55M, 4 GB DDR3 1333 Crucial
- Intel Core i5 2500/2500K and i7 2600/2600K, Intel DH67GD, 4 GB DDR3 1333 Crucial
- AMD Phenom II X4 975, Gigabyte 890GPA-UD3H, 4 GB DDR3 1333 Crucial
Far Cry 2For our games tests, we used three levels of different performance:
- 1280 x 720 low
- 1280 x 720 medium
- 1680 x 1050 medium
In Far Cry 2, the graphics modes used correspond to the low and medium modes on offer in the game. These modes use DirectX 9 exclusively.
The results with the Core i3 2100 HD 2000 are more or less similar to those with the i5 2500, except at the lowest quality mode where the processor puts the brakes on slightly. Coupled with the Radeon HD 5450, the trend is confirmed, the processor has an impact on performance in spite of the very modest graphics settings.
Crysis WarheadThe Low and Medium modes correspond to the ‘mainstream’ and ‘gamer’ modes in Crysis Warhead:
Performance levels between the Core i5 and Core i3 are closer here but the same thing is happening in terms of the CPU.
CPU impact vs IGPIn our Sandy Bridge test, we tried to measure if there was any impact resulting from the fact that the last level cache (LLC) was shared between the IGP and the x86 cores.
To try and get a reading for this, we opted to take a reading of the graphics and processor performance at the same time using Cinebench + Tom Clancy’s H.A.W.X. together. The percentage indicates the number of threads used for Cinebench (1, 2, 4) and corresponds to the percentage of threads in comparison to the number of physical cores on the chip (200% corresponds to 4 threads on the Core i3 2100 which has two physical cores + HT). We looked at the development in performance in these three cases:
While you’d expect to see a slight dip in performance on a dual core as you’re cumulating processor and graphics load (2.9% in the case of the Core i3 2100 + HD 5450 for graphics performance), it should be limited. When the Core i3 2100 is on its own with its IGP, graphics performance drops by 23%.
- Core i3 2100 + Radeon HD 5450
- Core i3 2100
- Core i5 661
Hold the mouse over the graph to see indexed performance.
We are still waiting for Intel to get back to us to clarify this issue that we had already noted with the Core i7 2600K. To recap, we had noted, other than the dip in performance, a dip in the IGP turbo clock. We then also recorded a drop in the IGP energy consumption using hwinfo 32. In the case of the Core i3 2100, the energy consumption in H.A.W.X alone is around 6 watts on a fixed game scene (moving at altitude). Launching Cinebench then leads to an almost instantaneous dip in the GPU energy consumption, falling to around 5 and then 4 watts and going as low as 2.5 watts in some cases. When we stopped Cinebench there was an immediate increase in performance and clock and the energy consumption rose to 6 watts instantaneously. In a similar scenario with the Core i7 2600K, the dip was from around 10 watts in IGP load to 3 watts in IGP and CPU load.
With an equal number of threads (4), it’s interesting to note that we observed a 22% reduction in graphics perforamnce on the Core i7 2600K, which is almost identical to what we got here. If this isn’t directly due to sharing the LLC, this dip in performance may well be linked to the ring bus which links the LLC to the x86 cores and IGP.
3D Studio Max, Cinema 4D
3D Studio Max 2010
We begin with the famous image rendering software, now in its x64 and 2010 version. The test scene used is from SPECapc for 3ds max 9 (space_flyby_mentalray) which employs the MentalRay rendering machine.
In 3d studio max, the Core i3 2100 at 3.1 GHz gives a 10.2% gain on the Core i3 540 at 3.06 GHz. This puts it between the Core i5-650 and 660. Compared to the AMD offer, it’s on a par with the Athlon II X3 445 at the same clock, but AMD also has the Athlon II X4 645 at the same price and this second model is notably faster.
Cinema 4D R11
The rendering software Maxon is well known in the overclocker community through Cinebench, which allows you to compare processor performance easily. We use version R11 of C4D in 64 bit mode with the scene from Cinebench R10 rendered at a higher resolution so as to prolong rendering time.
Cinema 4D gives an advantage of 12.4% to the Core i3-2100 over the Core i3-540. This puts the new CPU on a par with the Core i5-660. It’s much better placed in comparison to the AMD offer as it falls between the Athlon II X4 630 and 635. The 645 is however still out of reach.
MinGW / GCC
This is an applied test with the compilation of MAME source code using GCC under the MinGW development environment. We are now using version 5.1.4 of MinGW and compile the source code of Mame 0.133.
The Core i3 2100 is 11.9% faster here than the Core i3-540 in spite of being clocked only 1% higher. This puts it between the Core i5-650 and 660. The Athlon II X4 645 is however faster once again, with the Core i3 on a level with the Athlon II X4 630.
We’re using the 64-bit 3.9 version of WinRAR that introduces new multithreading optimisations to compress a group of files.
Putting the IMC on the die has an enormous impact on performance in WinRAR as the gain in comparison to the previous generation i3 is 32.6%! This gives the Core i3-2100 an excellent place in the standings – it isn’t outdone by the Core i5 LGA 1155.
H.264: Avidemux, MainConcept
Avidemux + x264
Our test videos use H.264 encoding exclusively. To start with, we use Avidemux version 2.5.2, which improves performance beyond 4 threads compared to version 2.5.1, to compress a 1920x1080 HD video file via the x264 codec at intermediary quality.
In Avidemux, the Core i3 2100 is 14.8% faster than the Core i3-540. This makes it only a little slower than the Core i5-670, but while this is enough to put it ahead of the Athlon II X3s, the Athlon II X4s (quads) are out front.
MainConcept Reference + H.264/AVC Pro
For this second H.264 encoding we use MainConcept Reference and its H.264/AVC Pro codec on “High”, still with the same video.
With a gain of 16.7% in comparison to the previous generation, the Core i3-2100 is on a par with the Core i5-670. It is only a hairsbreadth behind the Athlon II X4 630, but the 645 is still out of reach.
After Effects CS4, Nuendo 4.3
After Effects CS4
We’re using a new composition using various effects so as to render 3D animation. Multiprocecessing is activated so as to make the most of the available number of cores.
Intel processors dominate the AMD CPUs in this test, with the Core i3-2100 15.6% faster than the previous generation. It falls between the Core i5-660 and 670. Thanks to its 4 cores, the Athlon II X4s still holds their own and the 640 equals the Core i3-2100.
Here’s version 4 of Nuendo, with the latest patch 4.3, all at 64 bits. A new music project using various native plugins as well as 2 HalionOne virtual instruments was exported as a wav file (thanks to Draculax).
In Nuendo there’s a 14% gain in comparison to the Core i3-540. The Core i3-2100 is placed between the Core i5-660 and 670 and manages to outdo all the AMD processors in these tables.
Crysis & Arma 2
With patch 1.2, Crysis has a very heavy CPU bench (to be found in the Bin32/Bin64 directory). The test was carried out at high settings, but at a res of 800x600 so as to limit dependence on the graphics card.
Like in WinRAR, putting the memory controller on the same die as the processor has a big impact on performance in Crysis, giving the Core i3-2100 a 23% advantage over the Core i3-540. The Core i3-2100 falls between the Core i5 LGA1156 and LGA1155, leaving the AMD range far behind.
Arma 2 is configured with all settings at a max including max visibility (10 km), which brings the configurations to their knees. Resolution stays at 800x600 to stop the graphics card levelling performance. To gauge performance we measure the framerate during a well-defined movement after having loaded a saved game.
The Core i3-2100 is very much at ease in Arma 2 at no less than 33.3% up on the i3-540. This time, the Intel quad cores retain their advantage, only a slight one for the LGA 1155 versions, like some Phenom II X4s. The Athlon II X2s, X3s and even X4s are however trailing.
GTA 4 & Anno 1404
Grand Theft Auto IV
GTA IV is included in the protocol for its weight and multi-threading optimisations. Once again all the settings were pushed to a maximum, with the exception of the textures so as not to exceed available video memory, all at a res of 800x600. We use the built-in benchmark but on a scene chosen by us for more weight than the default.
In this third game, the Core i3-2100 is once again very impressive, with a framerate up by 26.7% on the i3-540. The Intel quad cores are a little faster, as are the highest end Phenom II X4s. The Athlon IIs can’t compete here.
Anno 1404 is a strategy game tested at max settings but with resolution still at 800x600. We use a saved game with a city of 46,600 inhabitants that we partly visualize from a distance.
Anno 1404 performance is much improved with the new Core i3, with a 32.8% gain on the previous generation. The 2100 is therefore by far the highest performance Intel dual core, but it still trails the quad cores. However, here the Core i3-2100 is in front of the Phenom II X4 975.
Hold the mouse over the graph to classify the CPUs by results.
AverageAlthough individual app results are worth looking at, especially for CPUs with a high number of cores, we calculated a performance index based on all tests with the same weight for each test. We attributed an index of 100 to the Intel Core 2 Q8200.
Thanks to the architectural improvements of the Sandy Bridge execution cores and putting the memory controller on the same die, the Core i3 2100 at just 3.1 GHz manages to outdo the Core i5-680 at 3.6 Ghz (up to 3.9 GHz in Turbo). At equal clocks, performance is up by around 20% on the previous generation, with the gain varying between 10.2% and 33.3% depending on the application.
At the same price, the i3-2000 is 16.5% faster than the i3-550 and 9% in front of the Athlon II X4 645. Of course, the average hides variations in different applications, with the Athlon II X4 645 taking the lead in applications that can exploit its four cores without loading the memory sub-system too much (3ds, Cinema 4D, MinGW, AVidemux, Mainconcept Reference and After Effects).
ConclusionThe LGA 1155 entry level dual core range is very attractive indeed. The memory controller, the IGP and the PCI Express are back on the same die, which allows the CPU to benefit from reduced latency during memory accesses, at the same time as reducing energy consumption, with everything manufactured at 32nm. In practice performances are up by an average of 20% at equal clocks on the 32nm LGA 1156 Core i3/i5s.
There is therefore a much bigger advantage than when you compare the 45nm LGA 1156 quad cores with the 32nm LGA 1155 quad cores (13.2% without HT or 11.3% with). In addition to the architectural improvements inherent to Sandy Bridge, the Core i3-2100 also benefits from the fact that the memory controller is once again on the same die as the CPU.
This gives a very significant gain in applications that are sensitive to reduced memory latency, with a gain of up to 33% in WinRAR, Crysis and Anno 1404 for a Core i3-2100 at 3.1 GHz in comparison to a Core i3-540 at 3.06 GHz. On average the Core i3-2100, priced at $117, is up on the Core i5-680, priced at $294!
The Athlon II X4s are not however completely pushed aside by the Core i3-2100. Version 645, which is at the same price, is 29.9% faster in 3ds, 7.8% faster in Cinema 4D, 10.7% faster in MinGW, 21% faster in Avidemux, 10% faster in MainConcept Reference and even 3.2% faster in After Effects, which is however favourable to Intel CPUs. This is of course because the Athlon II X4 645 has four physical cores, against 2 on the i3-2100, combined with the lesser impact of the memory & cache latency in these applications. On the other hand the i3-2100 is 48.8% faster in WinRAR, 29.9% in Nuendo, 58% in Crysis, 18.8% in Arma 2, 17.7% in GTA IV and 34.4% in Anno 1404.
You therefore need to make your choice in accordance with the end usage of the machine. If it’s going to be used mainly in highly multithreaded applications, the Athlon II X4 will have the advantage. On the other hand, the Core i3-2100 is superior in gaming applications. The Core i3 however clearly has the advantage in terms of energy consumption and heat dissipation and the same goes for the upgrade as the fastest LGA 1155 CPU (the Core i7-2600K) is up on the Phenom II X6 1100T.
Remember the Core i3-2100 IGP is also usable on the H67 platform. It’s an HD Graphics 2000 and not a 3000, as on the Core i3 Mobiles, but it shouldn’t be forgotten that there’s just a 3 MB cache which would have limited the peformance of additional units in any case. With its reduced temperature levels, the Core i3 could well be the ideal CPU for HTPCs, as long as the non-fuctional 23 Hz issue is quickly fixed.
In fact, the only real problem with this new range of dual core processors, is the fact that they can’t really be overclocked at all. Without Turbo, they don’t even have the 500 MHz margin proffered by Intel on the Core i5s/i7s, and no ‘K’ version is currently in the pipeline. This means we’ll have to make do with bus overclocking, which amounts to around 5%. This is a real shame as given the rather low initial clock and the results obtained on the i5s/i7s, a margin of 33 to 50% wouldn’t be surprising!
Does this mean you should avoid the LGA 1155 Core i3s? It’ll be hard to, given their other qualities!
Copyright © 1997-2013 BeHardware. All rights reserved.