Which of the following statements about artificial intelligence is wrong?
- A . Machine learning is the core of artificial intelligence
- B . The development trend of artificial intelligence technology is the open source of technology platforms, the development of special intelligence to general intelligence, and the development of intelligent perception to intelligent cognition.
- C . Artificial intelligence is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.
- D . Artificial intelligence is intelligence with humans as carriers
Which of the following is not a server virtualization software?
- A . FusionCompute
- B . XenServer
- C . FusionAccess
- D . VMware ESXi
Which of the following is not classified according to the network card structure type?
- A . Flexible card
- B . Mezz Card
- C . IlB network card
- D . Integrated network card LOM
After the ARMv7 architecture, the name of ARM’s processor architecture was changed to Cortex. Which ARM architecture does Huawei’s self-developed mobile processor chip Kirin use?
- A . Cortex-A
- B . Cortex-M
- C . Cortex-R
In Huawei Azure Stack hybrid cloud architecture, which of the following components is not a hybrid cloud infrastructure certified by Microsoft?
- A . FusionDirector Management
- B . FusionServer 2288H V5 2U2S rack server
- C . 10GE/25GE unified switching network
- D . Based on Huawei’s INVMe SSD hyper-converged storage
Which of the following semi-custom chips solves the shortcomings of custom circuits and overcomes the shortcomings of the limited number of gate circuits of the original programmable device?
- A . FPGA
- B . CPU
- C . GPU
- D . ASIC
Which of the following is not a component of the CPU processor?
- A . Register
- B . Logic Operation Unit
- C . Process
- D . Control unit
Server clusters are highly scalable. As demand and load grow, which of the following methods can be used to increase cluster performance?
- A . Increase the network bandwidth in the cluster
- B . Increase the memory capacity of a single server in the cluster
- C . Add a new cluster server to the cluster system
- D . Replace the hard disk of a single server in the cluster with a high IO solid-state drive
A memory module has the words "PC3L" on its label. What does the "L" in "PC3L" stand for?
- A . The memory module type is UDIMM
- B . Its rated power is 30W
- C . Its rated voltage is 1.35V
- D . The bit width of its memory chip is 64*3bit
Memory modules can be divided into UDIMM, RDIMM, and LRDIMM. Which of the following statements about these three types of memory modules is incorrect?
- A . LRDIMM breaks the maximum 16-rank limit for each channel and can increase the overall memory capacity of the system
- B . UDIMM is an unbuffered dual-channel memory module technology, which means that the address and control signals output by the controller directly reach the DIMM
- C . RDIMM means that the address and control signals output by the controller are output to the DRAM chip after passing through the Register, and the clock signal output by the controller reaches each DRAM chip after passing through the PLL.
- D . In LRDIMM, data and address control signals pass through the Register chip, which is suitable for large-capacity scenarios
Which of the following is a wrong understanding of the database cluster model?
- A . The two key points of a database cluster are: ① having two or more database server clusters, and ② being able to provide transparent services.
- B . Synchronize the database. After the database client issues a data update request, the node that receives the request will copy and transmit the updated data to other nodes in the cluster while returning the result to the client.
- C . Database clusters include two models: synchronous clusters and asynchronous clusters
- D . Asynchronous database: After the database client issues a data update request, the node that receives the request immediately returns the result to the client, and the updated data will be copied and transmitted to other nodes in the cluster at a certain time.
Huawei’s intelligent edge architecture has full-stack three-dimensional capabilities. Which of the following hardware level in the intelligent edge architecture does not belong to Huawei products?
- A . Smart Server G2500
- B . Atlas 300, a high-performance accelerator card based on Al processor
- C . Smart Station Atlas 500
- D . InfiniBand network card
Intelligent Platform Management Interface (IPMI) is an open standard hardware management interface specification that defines a specific method for () management subsystems to communicate.
- A . Combination
- B . Embedded
- C . Tiled
- D . Interface
BIOS is the basic input and output system, which provides the lowest level and most direct hardware control program for the computer.
- A . True
- B . False
Stateless computing means that the physical properties of computing servers can be remotely configured, inherited, copied and migrated, thereby automating server deployment, replacement and business migration without the need for reconfiguration, making computing "stateless".
- A . True
- B . False
As time goes by and technology advances, network cards have wider bandwidth and shorter latency.
- A . True
- B . False
In 2017, Ke Jie and AlphaGo’s man-machine battle 2.0 attracted countless people’s attention. The chip used by Google is an ASIC chip called TPU.
- A . True
- B . False
There are two main types of physical interfaces of network cards: electrical ports and optical ports.
- A . True
- B . False
Under the PCle2.0 standard, the available bandwidth of the X16 channel is ().
- A . 512 MB/s (duplex)
- B . 2.0 GB/s (duplex)
- C . 4.0 GB/s (duplex)
- D . 8.0 GB/s (duplex)
Scale-up refers to increasing () capacity of a server. It is applicable to financial transactions, weather analysis, scientific research and other fields.
- A . Single performance
- B . Single density
- C . Multiple density
- D . Multiple performance
Software is a series of computer data and instructions organized in a specific order. Which of the following statements about software is wrong?
- A . Middleware is software that provides connectivity between system software and application software.
- B . System software is responsible for managing various independent hardware in the computer system, providing a platform for running application software, and adapting to different application software.
- C . Application software is software developed for a specific purpose
- D . Computer software is divided into system software, application software and middleware between the two
Which of the following statements about HPC high performance computing is wrong?
- A . Early HPC was based on custom computers and was dedicated to specific fields such as basic scientific research and national defense strategy. With the development of technology, most HPC is now built on the cluster architecture of X86 servers.
- B . HPC stands for High Performance Computing, which refers to a single computer system that uses many processors or a computing system and environment that uses multiple computer clusters for computing purposes.
- C . HPC stands for supercomputing, which uses computers to research, design products and support complex decision-making
- D . HPC computing clusters are mainly composed of computing servers, shared storage, high-speed interconnection devices, and supporting management software, platform software, and application software.
Which of the following statements about the value of stateless computing is incorrect?
- A . Reduce the cost or cost of computing systems
- B . Solve the intelligent deployment, seamless replacement and migration of computing nodes
- C . Reduce the complexity and workload of maintenance
- D . Realize data link virtualization and network virtualization
GPU is widely used in embedded systems, mobile phones, personal computers, workstations and electronic gaming solutions. Which of the following descriptions about GPU is incorrect?
- A . GPU is mainly good at parallel computing such as image processing, and its processor system is SIMD (Single Instruction Multiple Data) type
- B . GPU is a general-purpose processor that takes into account both computing and control. Its processor system is SISD (Single Instruction Single Data) type
- C . GPU is very efficient for image and graphics processing. This is because GPU is designed with a highly parallel architecture, which makes it more advantageous than general-purpose processor CPU in large data block parallel processing algorithms.
- D . GPUs are made up of thousands of smaller, more efficient cores designed to handle multiple tasks simultaneously
Which of the following statements about RISC instruction set architecture is incorrect?
- A . RISC assembly language programs generally require a large amount of memory space, and the programs are complex and difficult to design when implementing special functions.
- B . RISC CPU contains a rich set of circuit units, so it has strong functions, large area and high power consumption.
- C . When implementing special functions on RISC machines, the efficiency may be low
- D . RISC designers focus on frequently used instructions and try to make them simple and efficient.
Which of the following models does not belong to Huawei’s SSD control chip?
- A . Hi1822E
- B . Hi1822
- C . Hi1812E
- D . Hi1811
Which of the following describes the BIOS operation process correctly?
- A . Power on -> Hardware initialization -> CPU loads BIOS -> Loads boot items -> POST boot hardware self-test -> Processes system instructions
- B . Power on -> Hardware initialization – > CPU loads BIOS -> POST hardware self-test -> Load boot items -> Process system instructions
- C . Power on -> CPU loads BIOS -> Hardware initialization -> Loads boot items -> POST boot hardware self-test -> Processes system instructions
- D . Power on -> CPU loads BIOS – > Hardware initialization -> POST hardware self-test -> Loads boot items -> Processes system instructions
Which of the following has a total of 64 PIN?
- A . 16x
- B . 1x
- C . 8x
- D . 4x
Which of the following is not a common transmission rate of a network card?
- A . 10Gb
- B . 5Gb
- C . 1Gb
- D . 25Gb
The CPU frequency includes main frequency, external frequency, bus frequency and frequency multiplier. Which of the following statements about frequency multiplier is wrong?
- A . PC overclocking usually involves adjusting the CPU multiplier without locking it. Adjusting the multiplier is much more stable than adjusting the external frequency.
- B . The data transmission speed between the CPU and the system is limited. If you blindly pursue a high main frequency and obtain a high frequency CPU, there will be an obvious "bottleneck" effect.
- C . Under the same external frequency, the higher the multiplier, the higher the CPU frequency, without limitation.
- D . The frequency multiplier refers to the relative ratio between the CPU main frequency and the external frequency.
Hard disks can be divided into mechanical hard disks, solid-state hard disks and hybrid hard disks according to storage media. Which of the following is the storage medium of solid-state hard disks?
- A . Magnetic head
- B . Disk
- C . Flash chip
- D . Integrated Circuit
Which of the following is not a major advantage of cluster systems?
- A . High performance
- B . High Availability
- C . High compatibility
- D . High scalability
Which of the following descriptions about server operating systems Windows and Linux is incorrect?
- A . Windows Server system has high requirements on server hardware and good stability
- B . Windows Server system is suitable for mid-range and low-end servers
- C . The Linux system has a complete set of permission mechanisms, with high security and stability
- D . Linux system is suitable for mid-range and high-end servers
Which of the following belongs to the upstream of the computing industry chain?
- A . Blockchain
- B . Chip Manufacturing
- C . Big Data
- D . Cloud computing
Scale out refers to splitting the work tasks among multiple servers for processing by which of the following?
- A . Distributed architecture
- B . Collective Architecture
- C . Composite Architecture
- D . Cluster architecture
The essence of intelligent computing includes breaking the boundaries of CPU, server and data center, and creating a full-stack scenario intelligent solution covering cloud, edge and end around a series of chips such as "computing, management, AI, storage and transmission".
- A . True
- B . False
A cluster is a group of service entities that work together to improve efficiency by increasing the number of tasks executed per unit time.
- A . True
- B . False
Server hardware mainly includes CPU, motherboard, memory, hard disk, PCle, chassis, power supply, fan, etc. For Huawei RH12288 v5, all components that support hot-swap include CPU, memory, and PCle expansion card.
- A . True
- B . False
General-purpose servers are based on PC architecture, using Intel or other processor chips compatible with the X86 instruction set and Linux or Windows operating systems. They are cheap, have good compatibility, poor stability, and not very secure. They are mainly used in small and medium-sized enterprises and non-critical businesses.
- A . True
- B . False
FPGA is a field programmable gate array, which is a programmable ASIC that can be customized to make targeted algorithm designs
- A . True
- B . False
Which of the following options is a common RAID type for servers?
- A . RAIDO/RAID1/RAID3/RAID5/RAID6/RAID10/RAID1E/RAID5O/RAID60
- B . RAIDO/RAID1/RAID3/RAID5/RAID6/RAID10/RAID50/RAID60
- C . RAIDO/RAID1 /RAID5/RAID6/RAID10/RAID 1E/RAID5O/RAID6O
- D . RAIDO/RAID1/RAID5/RAID6/RAID 10/RAID5O/RAID60
() is customer-oriented logic layer software, such as ERP, CRM, HR and other application software.
- A . Database Software
- B . Business application layer software
- C . Middleware
Common processor architectures include RISC (Reduced Instruction Set), CISC (Complex Instruction Set), and EPIC (Explicitly Parallel Instruction Set). The most widely used X86 architecture belongs to which of the following architectures?
- A . EPIC
- B . CISC
- C . RISC
- D . MIPS
Which of the following descriptions of the Apache + Tomcat cluster model is incorrect?
- A . Apache R-proxy means that when one of the Tomcats stops running, Apache will still forward the request, resulting in a 502 gateway error. However, this problem will not exist as long as the server is restarted.
- B . The disadvantage of Apache R – proxy is that when the stopped Tomcat server is started again, Apache cannot detect it and still will not forward the request.
- C . In Apache mod k mode, Apache will automatically detect the stopped Tomcat server and stop sending requests to it.
- D . Apache R-proxy) method has the advantage that it can be set up to allow ache to handle static web pages and let Tomest handle dynamic requests such as ISP and service.
Which of the following understandings about RAID10 is wrong?
- A . The first level is a RAID 1 mirror pair, and the second level is RAID 0
- B . RAID10 hard disk utilization is 50%
- C . RAID10 is a RAID level that combines mirroring and striping
- D . RAID10 has consistent read and write performance
What is the series of equipment developed by Huawei for artificial intelligence application computing platform?
- A . KunLun Series
- B . Atlas Series
- C . E9000 Series
- D . TaiShan Series
Which of the following statements about special computing is incorrect?
- A . Dedicated computing platforms mainly refer to mainframes and minicomputers.
- B . Dedicated computing platforms mostly use RISC architecture
- C . Dedicated computing platforms are expensive and have few applications. Only a few companies and a few applications can use computers.
- D . Huawei’s KunLun minicomputer belongs to a closed computing architecture
Which of the following is not a typical application scenario of GPU?
- A . General computing, completing complex logical operations
- B . Video image processing scenarios
- C . HPC hybrid application high throughput scenario
- D . HPC high-performance acceleration scenario
Heterogeneous computing is a new computing model after single-core computing and multi-core parallel computing. Which of the following background descriptions about heterogeneous computing is incorrect?
- A . Heterogeneous computing maximizes the performance of each processor and assigns the most suitable tasks to the most skilled computing unit.
- B . NUMA is a mode of heterogeneous computing
- C . Moore’s Law is getting closer to the physical limit, and it is becoming increasingly difficult and costly to improve single-core performance.
- D . The performance improvement brought by multi-core parallelism is limited. As the degree of parallelism increases, the problems of heat dissipation and energy consumption become increasingly prominent.
The database system generally consists of database, database management system, () and database users.
- A . Business application layer system
- B . Application System
- C . Data logic system