- 新着としてマーク
- ブックマーク
- 購読
- ミュート
- RSS フィードを購読する
- ハイライト
- 印刷
- 不適切なコンテンツを報告
Hi
First of all, I'm not a gamer. I'm a programmer and I need a fast system to compile and run programs written in Visual Studio in C++ and C# on Windows 7.
According to Amazon, the price of a six-core CPU (4930k) is 545$ and the price of the fastest quad-core (3770k) is 327$. It means six-core is 60% more expensive than the quad.
How faster is the six-core ?
Do you think this worth the excessive money ?
コピーされたリンク
- 新着としてマーク
- ブックマーク
- 購読
- ミュート
- RSS フィードを購読する
- ハイライト
- 印刷
- 不適切なコンテンツを報告
In looking at the specifications
Core i7 4930K Ivy Bridge, 12MB cache 3.4GHz
Core i7 3770K Ivy Bridge, 8MB cache, 3.5GHz
As to which is better for you, this depends on what you do and as to how the cache size may affect the performance.
For compile-debug-compile-debug... I do not think you will see much of a difference... unless your code has a large nesting of class/struct/templates. And in this case, the large cache may improve your compile time.
Also, compile time, for general purpose application development, is not affected too much after 2 or 4 cores. An exception to this would be a solution with many projects that are not dependent upon each other. IOW a Solution with independent libraries (these can compile independently).
Now then,
Once you compile your program, and run it in Release Build, you have to assess as to if the extra cores would make a difference.
Jim Dempsey
- 新着としてマーク
- ブックマーク
- 購読
- ミュート
- RSS フィードを購読する
- ハイライト
- 印刷
- 不適切なコンテンツを報告
That depends what applications are you going to run on these machines.If you only deal with compling everyday ,i think quad-core would be enough ,parallel-compilation is limited to utilize all the cores. Maybe you are writing some multi-threading c++/c# applications which executes concurrently a lot at runtime?
Thanks ,Qiaomin
- 新着としてマーク
- ブックマーク
- 購読
- ミュート
- RSS フィードを購読する
- ハイライト
- 印刷
- 不適切なコンテンツを報告
It usually depends on the type of application beign developed by you.For example when you take into account heavy floating point code then additional cores with it shared vector execution unit(per core) could probably speed up the execution of such a code.When this is nicely parallelized.
- 新着としてマーク
- ブックマーク
- 購読
- ミュート
- RSS フィードを購読する
- ハイライト
- 印刷
- 不適切なコンテンツを報告
I would say that you should consider :
1 - Is your application parallelized and able to run with 12 threads ?
2 - Your application needs a big L3 cache ?
If the answer to the first question is yes you may consider the six core. If the answer to both questions is positive you can be sure about six core.
If your application is not able to run with 8 thread or more, I would say that a four core is the best choice.
There are other points of view: your clients and how fast are other competitors in your market.
