I’m working on some server specifications for a large company that will be used mostly for high transaction systems and database servers. I noticed that there are some new processors on higher end servers. I will be going with the Compaq ML570 and DL580 line. I’ve read all the marketing from Intel, and they swear that the new Xeon processors outperform the old Xeon processors in all categories. My one concern is the amount of cache on board the newer Xeon processors. I’m not familiar with the details of exactly how processors use cache and exactly how it helps performance.
The older model uses 4 900MHz Xeon processors with 2mb L2 cache
The new model uses 4 1400MHz Xeon processors with 256k L2 cache and 512k L3 cache. The new model also boasts hyperthreading technology.
In my experience, using processors with more cache really helps CPU intensive applications running on Windows NT/2000. I’d be interested in hearing some opinions, and if possible, finding some actual independent benchmarks of these two configurations.
It would really depend on the price, but I have worked with the Compaq DL 580 recently, a dual 1400Mhz Xeon, and it is quite fast. I compared the responsiveness with another client that has Quad PIII Xeon 733, and well, I'd go with the dual P4 Xeons any day. Cache does matter, but only if you compare same speed CPU's. 256k L2 at 1400Mhz and 512k L3 at 400Mhz is better than 2MB L2 at 900Mhz in my opinion.
If price is not a problem, I would definitely go with the P4 Xeons. The 4 processors with Hyperthreading should help out at least 10% to 20% (if not more) with DB operations, with regards to a comparable system that doesn't use Hyperthreading.
XEON P3 built core with beef up 2 mb of L2 vs a XEON MP P4 built + L3.This old stuff Gallatin will be release soon this Quarter but price is high 3000/2000/1500 but more reliable
Well, I know what happened in the PC market. The PIII 1000EB was rougly the same performance as the P4 1.4GHz. These are the same cores used for the Xeon 900 and the Xeon 1400.
Now in the PC market, the 900 moderately underperformed the 933 simply because it ran at a slower bus speed. The PIII Xeon has the same limitation of the PIII 900, a 100MHz bus.
So the 900 Xeon should have about 15% lower performance than the 1400 Xeon, if they had the same amount of cache. But because the 900 2MB has 8x as much cache as the 1400, it will probably perform slightly better in such applications that make good use of cache.
I wouldn't upgrade.
<font color=blue>You're posting in a forum with class. It may be third class, but it's still class!</font color=blue>
You are about to answer a thread that has been inactive for more than 6 months. If you still wish to proceed, please ensure that your posting is original and does not duplicate or overlap any prior responses to this thread.