Re: Which processors run SIMD to take advantage of Oracle 12c in memory option parallel processing

On Monday, August 25, 2014 12:17:32 AM UTC-4, Mladen Gogala wrote:
> On Sun, 24 Aug 2014 20:04:25 -0700, zigzagdna wrote: > My company has linux servers which use AMD Opteron(TM) Processor 6276 > We are running Red Hat enterprise Linux 6.x. Do these processor support > SIMD which can be used with Oracle 12c in memory database option to > access column store very fast by doing parallel processing. I did google > search, but could not find this information; so hoping experts here can > answer my question. Zigzagdna, you have confused things completely, not for the first time. Oracle 12c in-memory doesn't option require SIMD, which is also known as "vector processing". Vector processing is very handy when dealing with large floating point matrices but plays no role in resolving queries by using in-memory option. Oracle in-memory option uses smart hashing algorithms and statistic methods like Bloom filters which are not aided by the vector processing. And RH Linux 6.x can use any Opteron CPU. In other words, the answer to your question is "42". -- Mladen Gogala The Oracle Whisperer http://mgogala.byethost5.com

I know in memory option does not reuqires SIMD, but it can benefit from it.  Larry's video refers to SUN's ultra(?) Sparc processors; but I do not hvae them, thta's why I was aasking whether AMD Opteron(TM) Processor 6276 has this capability.

Your statment that SIMD does not play any part in in memory option is not correct. Please refer to Maria's blog

. Use SIMD to apply filter predicates

For the IMCU that need to be scanned Oracle takes advantage of SIMD vector processing (Single Instruction processing Multiple Data values). Instead of evaluating each entry in the column one at a time, SIMD vector processing allows a set of column values to be evaluated together in a single CPU instruction. The column format used in the IM column store has been specifically designed to maximize the number of column entries that can be loaded into the vector registers on the CPU and evaluated in a single CPU instruction. SIMD vector processing enables the Oracle Database In-Memory to scan billion of rows per second per core versus the millions of rows per second per core scan rate that can be achieved in the buffer cache. Received on Mon Aug 25 2014 - 13:53:07 CEST