Dedicated servers based on Intel Avoton processors

This year, Intel introduced the new Atom processors from the C2000 line, also known under the code name Avoton. We will describe in detail about new configurations and their technical features in this article.

Specifications

The 64-bit Intel Atom C2000 family of processors, also known under the codename Avoton, replaced the previous generation (codenamed Intel Centerton), which for a number of reasons turned out to be of little demand.

New processors are released on the 22-nanometer process. They include from 2 to 8 cores. We offer servers based on the older in the line of the eight-core Intel Atom C2758 processor (their configurations are described in detail below).
')
Processors Intel Atom C2000 family support the extraordinary execution of 64-bit commands (out-of-order), which allows to increase performance. Supported and technology Turbo Boost, through which the performance is automatically increased at peak loads. So, the 2.4 GHz clock frequency can be increased to 2.7 GHz with Turbo Boost.

The level of energy consumption of new processors is in the range of 5 to 20 watts.

Like previous models, the Intel Atom C2000 processor family has 16 PCI Express lanes and supports LPC, SPI, SMBus, and UART interfaces. Dual channel DDR controller works with DDRIII-1600 and DDRIIIL-1600 memory.

The new processors are based on the Silvermont microarchitecture, thanks to which a significant increase in performance has been achieved. Consider its features in more detail.

New architecture

The architecture of Intel Atom processors for a long time did not undergo major changes (except for the transition from 45-nanometer to 32-nanometer technological process). The emergence of the Silvermont microarchitecture has become a serious step forward: the processors of the new family are radically different from all previous “atom” models.

The new micro-architecture is based on the extraordinary execution of commands, thanks to which the performance has increased significantly. In this case, Intel is still based on the principle of combining certain instructions into a single instruction-micro-operation, which increases the efficiency of processing the corresponding x86 commands.

For processors with the Saltwell microarchitecture (Intel Centerton processors are based on it) the length of the computing pipeline is 16 steps. Since these processors are based on the paradigm of sequential execution of instructions, micro-operations must go through all the steps of the pipeline, even if they do not need cache access steps. As a result, the branch with the prediction error of the transition spends 13 cycles in vain.

In the Intel Silvermont architecture, micro-operations can bypass cache access steps and execute immediately in cases where the cache is not involved. Therefore erroneous
prediction spends only 10 cycles.

Many instructions that were slowly executed in the previous architecture were modified to reduce latency and increase throughput. A few clock cycles reduced floating point calculations. SIMD paired calculations are now performed in 4 cycles, not 9, as before. The number of instructions executed per clock has increased approximately 1.5 times.

The L2 cache in Intel Silvermont processor-based x is tightly integrated with the cores. The cores are combined into modules — pairs with a shared L2 cache of 1 MB in size. Separate cores, the L2 cache and the interface between the cores and the cache can be powered independently of each other. The cores of the same module can operate at different frequencies, but by default they work symmetrically.

The modules communicate with each other using the IDI interface (in-die interface implemented on the chip itself), which has independent read / write channels.

One of the distinguishing features of the Silvermont architecture is the rejection of HyperThreading technology. Improving the performance of many streaming applications is achieved by increasing the number of physical cores.

In order to improve performance in Silvermont-based processors, the instruction set has been significantly updated. There was also an acceleration of AES-NI encryption and Secure Key.

Virtualization hardware support is based on second-generation VT-x technology. The set of Extended Page Tables functions are supported, including the addressing of the Virtual Processor ID in the TLB buffer and the Unrestricted Guest mode, through which KVM users can access the executable code.

At the instruction and functionality level, the Intel Atom C2000 processor family is fully compatible with the Xeon processor family.

Performance

To give a more visual idea of ​​the performance of new processors, we present a table with test results.

Testing was done using the sysbench utility (CPU, Threads, Mutex, Memory columns). The smaller the resulting value, the better the performance.

The table also includes the results of the calculation of relative performance per core (column “Performance of 1 core”) and per core with frequency normalization (column “Architecture efficiency”). The higher the value, the better the result.

Model	CPU	Threads	Mutex	Memory	Performance 1 core	Architecture Efficiency	Kernels	Frequency, GHz
2x Xeon E5-2630	37.273	39.992	4.065	67.559	2.24	0.97	12	2.3
2x Xeon E5-2620	42.101	42.878	4.464	73.784	1.98	0.99	12	2
2x Xeon E5620	59.907	29.99	2.832	61.652	2.09	0.87	eight	2.4
2x Xeon L5630	71.034	34.059	3.348	68.583	1.76	0.83	eight	2.13
Xeon E3-1270	72.565	19.616	1.523	10.8	3.45	1.01	four	3.4
Xeon E3-1230	87.891	26.902	1.622	23.348	2.84	0.89	four	3.2
2x Xeon E5504	94.221	34.645	3.313	77.046	1.33	0.66	eight	2
Avoton C-2758	135.624	29.7066	5.25	78.895	0.92	0.38	eight	2.4
2x Xeon E5-2603	142.662	33.183	4.575	89.117	0.88	0.49	eight	1.8
Core2 Q8300	151.55	31.28	4.345	71.583	1.65	0.66	four	2.5
Core i3-2120	175.782	29.013	1.418	18.201	2.84	0.86	2	3.3
Core2 E8400	252.464	43.443	0.968	26.514	1.98	0.66	2	3
2x Xeon E5130	339.772	45.813	6.838	90.813	0.74	0.37	four	2
Atom d525	2159.999	149.234	8.177	210.218	0.23	0.13	2	1.8

The table shows that due to the improvement in the processor architecture, the computation efficiency has increased almost three times. The new “Atoms” have risen to about the level of Xeon processors of older generations, they almost fall short of the level of Core2 microarchitecture. The lag in the performance of the architecture from modern Xeon models is compensated by the high frequency of the cores and their number, which favorably affects the final performance.

The test results show that the new “Atoms” are capable of a much wider range of tasks compared to previous models. Today, it is safe to say that Atom processors are no longer a niche product and are striving in the low-middle segment of the server market. Low cost, low power consumption and ample opportunities undoubtedly make attractive to a wide range of potential users.

New configurations and their capabilities

We offer the following server configurations based on Intel Avoton processors:

• Intel Atom C2758 2.4GHz, 8GB DDR3, 2 × 1TB SATA - 3000 rubles per month;
• Intel Atom C2758 2.4GHz, 16GB DDR3, 2 × 240GB SSD - 4000 rubles per month.

New servers will undoubtedly attract many users because they are characterized by an excellent price-quality ratio.

The first configuration (Intel Atom C2758 2.4GHz, 8GB DDR3, 2 × 1TB SATA) is well suited for solving such tasks as serving static web pages, dedicated entry-level hosting, data delivery, etc.

Model c SSD can be used as a front-end server, as well as a caching server. New servers are available for order both in St. Petersburg and in Moscow.

Readers who can not leave comments here are invited to our blog .