Does full-speed 256 L2 cache reduce SMP efficiency?

Does full-speed 256 L2 cache reduce SMP efficiency?

Post by Steve Snyde » Wed, 16 Feb 2000 04:00:00



I have read that a full-speed 256KB L2 cache reduces SMP efficiency
relative to a half-speed 512KB cache.  For example, a P3/600e vs. a
P3/600.

Is this true?  Is there greater benefit in SMP to a larger L2 cache,
even if it is half as fast?

Thank you.

***** Steve Snyder *****

 
 
 

1. Q: Shared L2 cache on SMP system

I've purchased recently this specific motherboard (Tomcat IV Dual) with
2 P200/MMX, they're worth 398 BogoMips each (that's 796 BogoMips all
together !). But I don't know if this figure means anything at all.

Otherwise, the performance is not too bad at all.. :) Be a little more
specific about what you'd like to know about it.

--

Software Engineer      Brokat Systeme     hotline +49 1805 217171
                odi profanum vulgus et arceo
        PGP key ID 2B37D181 available on key servers

2. ppp woes.

3. Raise MAX_MP_BUSSES to 256 on SMP boxes

4. SCSI CD-RW: yamaha 4416s: HowTo REWrite CD-RW's?

5. SPiDER 256-CACHE and X.

6. Changing shells on Solaris 2.5 for non-root accounts

7. GUS + BusLogic BT-445S + STB Pegasus + DX2/66 + 256 kB cache = trouble

8. fork, pthread_create and the FP stack

9. 256 or 512 of PB cache?

10. Q: System w/ limited 256 kB cache, 16MB->32MB = slower linux.

11. Apache Proxy -- cache statistics (%full, cache-hit-rate)?

12. Linux 2.0.30, SMP, Efficiency, and Distributed Net/RC5

13. L2 Cache