SynQuacer E-Series Motherboard for 24-core ARM Server on ARM Cortex A53 Processor with 32 GB of RAM,
For many years, reduced instruction set (RISC) ARM processors have dominated the mobile device market. But they never managed to break into the data centers, where Intel and AMD with the x86 instruction set still dominate. Separate exotic solutions appear periodically, such as , but there are no serious proposals yet. In fact, it wasn't until this week.
AWS launched its own 64-core ARM processors in the cloud this week is a system-on-a-chip with an ARM Neoverse N1 core. The company claims that Graviton2 is much faster than previous generation ARM processors in EC2 A1 instances, and here are .
Infrastructure business is a comparison of numbers. In fact, customers of a data center or cloud service do not care what architecture the processors have. They care about the price/performance ratio. If work on ARM is cheaper than on x86, then they will be chosen.
Until recently, it was impossible to say unequivocally that computing on ARM would be more profitable than on x86. For example, a server 24-core ARM Cortex A53 is a model costing about $1000, which could raise a web server on Ubuntu, but was much inferior in performance to the x86 processor.
However, the amazing energy efficiency of ARM processors makes you look at them again and again. For example, SocioNext SC2A11 consumes only 5 watts. But electricity accounts for almost 20% of data center costs. If these chips show decent performance, then x86 won't stand a chance.
First Coming ARM: EC2 A1 Instances
At the end of 2018, AWS introduced on their own ARM processors. Certainly, this was a signal to the industry about potential changes in the market, but the results of the benchmarks were disappointing.
The table below shows EC2 A1 (ARM) and EC2 M5d.metal (x86) instances. The utility was used for testing. stress-ng:
stress-ng --metrics-brief --cache 16 --icache 16 --matrix 16 --cpu 16 --memcpy 16 --qsort 16 --dentry 16 --timer 16 -t 1m
As you can see, A1 performed worse in all tests, except for the cache. For most other indicators, ARM was inferior very much. This performance difference is greater than the 46% price difference between the A1 and M5. In other words, instances on x86 processors were still more profitable in terms of price / performance ratio:
Test
EC2 A1
EC2 M5d.metal
Difference
cache
1280
311
311,58%
icache
18209
34368
-47,02%
matrix
77932
252190
-69,10%
cpu
9336
24077
-61,22%
mempy
21085
111877
-81,15%
qsort
522
728
-28,30%
dentry
1389634
2770985
-49.85%
timer
4970125
15367075
-67,66%
Of course, microbenchmarks do not always show an objective picture. What matters is the difference in the actual performance of the application. But even here the picture was no better. Colleagues from Scylla compared a1.metal and m5.4xlarge instances with the same number of processors. In a standard NoSQL database read test in a single node configuration, the first showed 102 reads per second, and the second 000. In both cases, all available processors are used at 610%. This equates to a reduction in performance of about six times, which is not compensated by the lower price.
In addition, A1 instances run only on EBS without support for fast NVMe devices like other instances.
All in all, the A1 was a step in a new direction, but didn't live up to expectations for ARM.
Second Coming of ARM: EC2 M6 Instances
That all changed this week when AWS introduced a new class of ARM servers, as well as a number of instances on new processors. , including .
Comparison of these instances shows a completely different picture. In some tests, ARM performs better, and sometimes much better, than x86.
Here are the results of running the same stress testing command:
Test
EC2 M6g
EC2 M5d.metal
Difference
cache
218
311
-29,90%
icache
45887
34368
33,52%
matrix
453982
252190
80,02%
cpu
14694
24077
-38,97%
mempy
134711
111877
20,53%
qsort
943
728
29,53%
dentry
3088242
2770985
11,45%
timer
55515663
15367075
261,26%
That's a completely different story: the M6g is five times faster than the A1 when reading from a Scylla NoSQL database, and the new M6gd instances run fast NVMe drives.
ARM offensive on all fronts
The AWS Graviton2 processor is just one example of ARM being used in data centers. But the signals are coming from different directions. For example, on November 15, 2019, the American startup Nuvia .
The startup was founded by three lead engineers who were involved in the creation of processors at Apple and Google. They promise to develop processors for data centers that will compete with Intel and AMD.
On , Nuvia has designed a processor core from the ground up that can be built "on top" of the ARM architecture, but without obtaining an ARM license.
All this indicates that ARM processors are ready to conquer the server market. After all, we live in a post-PC era. Annual shipments of x86 have dropped nearly 10% since their peak in 2011, while RISC chips have skyrocketed to 20 billion. Today, 99% of the world's 32- and 64-bit processors are RISC.
Turing Award winners John Hennessy and David Patterson published an article in February 2019 . Here is what they write:
The market settled the dispute between RISC and CISC. While CISC won the later stages of the PC era, RISC is winning now that the post-PC era has arrived. No new ISAs have been created on CISC for decades. To our surprise, the general consensus on the best ISA principles for general purpose processors today still leans in favor of RISC, 35 years after its invention... In open source ecosystems, cleverly designed chips will convincingly demonstrate advances and thereby accelerate commercial adoption. The general purpose processor philosophy in these chips is likely to be RISC, which has stood the test of time. Expect the same rapid innovation as during the last golden age, but this time in terms of cost, energy and safety, not just performance.
“The next decade will see the Cambrian explosion of new computer architectures, which means exciting times for computer architects in academia and industry,” they conclude the article.
Source: habr.com