Tag Info

Hot answers tagged

22

Your problem doesn't have (almost) anything to do with the CPU. Transferring big files is usually fast, since it can be done with sequential I/O. Transferring lots of small files requires tons of horsepower on the storage side of things, since it requires random I/O. Low seek times, fast hard drives, lots of cache and a filesystem designed for huge number ...


15

With the HP ProLiant G5 and newer (G6, G7, Gen8) servers (e.g. DL380, ML370, etc. - Anything after the Intel 5400-series CPU's), it is possible to disable half of the cores available to the server. This is a BIOS switch labeled "Processor Core Disable" with options "All Processors Enabled" and "Disable One-Half of cores per Physical Processor." This is a ...


14

The command psrinfo -pv is the command you are looking for. It gives you the number of physical cpus plus the count of virtual processor per physical processor. For example on a V880 it looks like this: The physical processor has 1 virtual processor (0) UltraSPARC-III+ (portid 0 impl 0x15 ver 0x23 clock 900 MHz) The physical processor has 1 virtual ...


13

Should be very little difference in performance between two dual core CPUs and a single quad core CPU, if you are using current Intel Xeons. However, you might expect to see slightly better performance from the quad-core CPU if it is a "true" Quad core design, ala the new AMD Opterons or the new Xeon W5xxx (aka the Core i7 / Nehalem) series. ...


11

I don't see an idle CPU core in your output. Core 0 is 100% active servicing software interrupts, all the others are split between user/system and idle.


10

According to this information provided by Microsoft, it depends on the version of 7: http://windows.microsoft.com/en-US/windows7/products/system-requirements I would quote, but phone copy/paste with chrome is not cooperative. Supposedly any 32 bit 7 edition will support 32 cores. Any 64 bit 7 edition will support 256 cores. Further, the maximum ...


10

There are two factors you want to look at: Total aggregate performance: This is the measure of the total computing power of all the cores on the box. You can estimate this value by looking at the Passmark CPU rating for the CPU and multiplying by the number of physical CPUs in the system. Single-thread performance: This is the measure of how much computing ...


9

On libvirt 0.8.3, if you type: virsh capabilities | grep topology it will list the topology of the host: <topology sockets='1' cores='4' threads='1'/> The numbers refer to sockets, cores per socket, and threads per core. Add this line to the cpu entry in the xml file to allow windows to use all 4 cores, e.g.: <vcpu>4</vcpu> ...


8

No, they don't. Each and every blade is a distinct machine with it's own CPUs/IO ports/memory. They do connect to an (active or passive) backplane (which connects the verious IO options like ethernet etc.) and share the cooling and power equipment.


8

Technically SQL Server Express does not use one CPU, it uses one SOS Scheduler. Meaning it has only one active thread in the process at any moment (one worker has the one scheduler for himself, until it yields), ignoring some special threads like DAC. This SOS scheduler does not make any enforcing of the CPU to run on, that is left to the OS, so the SQL ...


8

I suspect the processors are being presented as single core processors in separate sockets. Windows 7 ultimate supports up to 2 socket systems, so it will use 2 processors. I don't know if you can configure KVM to present the processors as either a single quad core CPU or 2 dual core CPUs, which should resolve the problem.


8

JustinC already linked to the document I am referring to: Keep in mind that there is a physical CPU limit of 2 with Windows 7 Professional, Ultimate and Enterprise. You are planning to use 4 processors, this will not work.


7

Use '&' after a command to background a process, and 'wait' to wait for them to finish. Use '()' around the commands if you need to create a sub-shell. #!/bin/bash curl -s -o foo http://example.com/file1 && echo "done1" & curl -s -o bar http://example.com/file2 && echo "done2" & curl -s -o baz http://example.com/file3 ...


7

If there are any core-to-core communications then you'll find that generally a quad-core will be faster, sometimes significantly so, than two dual-cores. The reason is that the bandwidth and/or latency between cores on the same die can often be very much faster than any external connections. That said you don't mention specific processors so you could be ...


7

That CPU isn't idle, if that's what you're thinking. 100.0%si shows that it's spending its time on handling interrupts.


7

Your performance may have nothing to do with core-count. Yo could be I/O-constrained or just not have enough RAM on the system. What are the specifications of the underlying system hardware? E.g. CPU model, clock speed, RAM amount. Is the performance poor compared to another metric? Your expectations? Has the performance always been bad? In general, I ...


7

David showed you where the bazooka is, but if you use it you will almost certainly be shooting yourself in the foot (see my comment). As a general rule you should accept that you are not smarter than the task scheduler -- A lot of very bright people worked on it, and they did a really good job of making sure it crams the maximum number of computations into ...


7

No. GPU cores are (comparatively) very simple devices. They work great at massively parallel tasks, like rendering, encryption, and other math. They're basically horrible at logic, branching, and dependent operations (greatly simplified for the sake of argument). They also do not run the same code as a CPU. A task like serving web pages involves almost no ...


7

You could consider running VMWare ESXi on the hardware, and only present 4 processing cores to the guest. The "overhead" should be negotiable as there would be four free cores.


7

In my experience using a 32 and 64 bit variants of FreeBSD on the same hardware, I have seen a measurable (but not huge) benefit to running 64 bit. 64 bit has been around long enough that active projects don't have issues. The one area this is not necessarily true is for compiled commercial apps (doesn't sound like an issue for your workload). If you were ...


7

Set the lowest number of vCPUs your servers need to perform their function, don't over-allocate them or you could easily slow down your VMs.


6

I believe you just add vcpus=2 to the guest's config file.


6

The x86_64 architecture has a few other advantages than just the memory space. The most imporant one is the increased number of registers, which allows the compiler to generate more efficient code. It's unlikely to make a big difference in the performance of your system though.


6

If you want to be able to tune MySQL to utilize multiple cores, you need to upgrade to MySQL 5.5. According to MySQL's Whitepaper "What’s New in MySQL 5.5 Performance and Scalability" from December 2010: Control of Background I/O Threads – Users now have two new configuration parameters for all platforms, innodb_read_io_threads and ...


6

Based on your comment on MikeyB's answer you're trying to solve this the wrong way IMHO -- Both numactl and taskset will lock your process to a CPU, but they won't keep other processes off that CPU. If someone else is on that CPU when your process needs it you will have to wait. A better solution is to set your process' nice value to something that will ...


6

That depends. If the CPU utilization is from user, then the answer is most likely "yes". If the CPU utilization is from iowait, then you're wasting your money until you upgrade your storage system. If it's mostly from system, then you probably have a driver or hardware device that is taking up too many interrupt resources (network card, maybe?)


6

Taskset is for binding a process to one or more CPUs; essentially specifying where it can run at initial execution or while it's running. If using RHEL/CentOS on modern server equipment, numactl is recommended over taskset. Cpuset/cset is for CPU shielding and is a framework build around Linux cgroups. Cset was never popular on certain distributions (like ...


5

This is a complex theme, but do not forget one other element here - you also compare hardware of different generations. This is not ONLY a pure GHZ comparison, you may very well compare outdated opterons (8 cores only? - is that 2x4, that is a VERY slow core compare to todays) with a sandy bridge / Ivy bridge level Xeon. I would go with the Xeon, simply for ...


5

There is a lot more to your question than just cores and performance. How many concurrent users does your server need to support? Are you with one server and no redundancy? What is the acceptable downtime for your application? Have you done any benchmark of your development machine for this application to somewhat extrapolate performance? You may be ...


5

An E5620 CPU has 4 cores on the die. With Hyper-Threading turned on, that gives you 8 threads. Perhaps your machine has 2 physical E5620 CPUs? If it's not powered on, try opening the case and counting them. If it's powered up and you can log in, try this: cat /proc/cpuinfo | grep “physical id” | sort | uniq | wc -l 2 physical CPUs x 4 cores/CPU x 2 ...



Only top voted, non community-wiki answers of a minimum length are eligible