Whats the diference between new and old processors?

griffin81

Distinguished
Sep 5, 2015
6
0
18,510
I had a wonderful i5-2500, a procesor launched at 2011. And it has been working great until now. My motherboard has died and looks like I cannot find a replacement for that socket (1155), only refurbished **** and some H61 that doesn't works with my cpu.

So I have decided to replace the motherboard and the cpu. I'm looking for its natural replacement, the i5-6500, but looking at the specifications looks like is not much better than the one I had: http://ark.intel.com/compare/88184,52209

So how is that possible, in 5 years the only achievement is a higher bus speed, a slighty lower power consumption that I don't really care about, and support for 64Gb of RAM instead of 32GB (I have 8 so I don't care about this either)? And then the clock speed of the new processor is also lower than the old one?
 
Solution
There's a lot more going on in modern CPUs. Here's an article on Sandy Bridge's successor, Ivy Bridge:

http://www.anandtech.com/show/4830/intels-ivy-bridge-architecture-exposed

Sample:

The front end in Ivy Bridge is still 4-wide with support for fusion of both x86 instructions and decoded uOps. The uOp cache introduced in Sandy Bridge remains in Ivy with no major changes.

Some structures within the chip are now better optimized for single threaded execution. Hyper Threading requires a bunch of partitioning of internal structures (e.g. buffers/queues) to allow instructions from multiple threads to use those structures simultaneously. In Sandy Bridge, many of those structures are statically partitioned. If you have a buffer that...
There's a lot more going on in modern CPUs. Here's an article on Sandy Bridge's successor, Ivy Bridge:

http://www.anandtech.com/show/4830/intels-ivy-bridge-architecture-exposed

Sample:

The front end in Ivy Bridge is still 4-wide with support for fusion of both x86 instructions and decoded uOps. The uOp cache introduced in Sandy Bridge remains in Ivy with no major changes.

Some structures within the chip are now better optimized for single threaded execution. Hyper Threading requires a bunch of partitioning of internal structures (e.g. buffers/queues) to allow instructions from multiple threads to use those structures simultaneously. In Sandy Bridge, many of those structures are statically partitioned. If you have a buffer that can hold 20 entries, each thread gets up to 10 entries in the buffer. In the event of a single threaded workload, half of the buffer goes unused. Ivy Bridge reworks a number of these data structures to dynamically allocate resources to threads. Now if there's only a single thread active, these structures will dedicate all resources to servicing that thread. One such example is the DSB queue that serves the uOp cache mentioned above. There's a lookup mechanism for putting uOps into the cache. Those requests are placed into the DSB queue, which used to be split evenly between threads. In Ivy Bridge the DSB queue is allocated dynamically to one or both threads.

Dunno how technical you want to get, but at a really basic level, new CPUs do more for every clock cycle, and do it using less power. In notebooks and servers, where power consumption and efficiency are pretty much the most important aspects of a CPU, they're several times faster now than they were 5-6 years ago. On desktops where power consumption is basically an afterthought, the performance improvement has been a modest ~40% (give or take) in non-AVX workloads. Intel doubled the width of the AVX pipeline from Ivy Bridge to Haswell, so under that instruction set, things can be more than twice as fast.
 
Solution
what most people do not understand is that physical specs like clock speed, core, and cache have absolutely nothing to do with performance

older CPUs have less transistors, instructions per cycle IPC, and sometimes cache and less instructions on the CPU itself. the older CPUs also are less efficient. you can't look at the outside man. older CPUs aren't viable options anymore due to this. that's why older Xeon X5650 CPUs have to get a 90% overclock to be any effective.

http://www.cpu-world.com/Compare/415/Intel_Core_i5_i5-2500_vs_Intel_Core_i5_i5-6500.html
in this specific situation, the CPU you are trying to compare has little difference, because it is still a quite good line if CPUs an can handle. only in some apps it can perfrom faster, as the site says. regardless of synthetic benches, the site details good.

an option for you is to look on ebay for the older motherboards. people will be trying to rid of them for the newer CPUs.
 


I found an H61M-k from asus and looks like only supports 77W processors while mine have 95W. The anti surge protection was rebooting the system randomly while I played videogames.
 
I see ton's of LGA1155 socket motherboards available. I selected your CPU on PCPartpicker.com, and then went to the select motherboard page and this is what I see:

http://pcpartpicker.com/products/motherboard/

If the link doesn't work properly, just select your CPU and then go to motherboards.
 


Thanks,

Your answer also helped me a lot to understand it. Wish I could choose both as an answer.

 


Ouch, I forgot to say that I live on Spain, and it's hard to find foreign electronics shops that ships here.
 
Well, even if you can't get an electronics shop outside of Spain to ship there, now you have a list of all of the motherboards that would work with your old CPU, so you could shop around locally, or if there is an online shop in Spain. (I "think" I selected the Compatibility Check in PCPartpicker.com, but if I didn't you can go there and do that).