BM Seer Unofficial thoughts from an anonymous Sun employee

World Record Lotus Domino R6iNotes Sun SPARC Enterprise 5220

Tuesday Oct 09, 2007

The Sun SPARC Enterprise T5220 (one UltraSPARC T2), using Lotus Domino 7.0.1 mail server delivered industry-leading results with  best per-socket performance and best power-performance for the Lotus[R] R6iNotes on Domino mail server benchmark. The Sun SPARC Enterprise T5220 has better $/performance than all IBM and HP servers on this benchmark.

The Sun SPARC Enterprise T5220 server achieved 43,000 users on the Lotus Domino 7.0.1 Messaging server running the HTTP based R6iNotes benchmark.

The Sun SPARC Enterprise T5220 is 35% faster than the 4-core IBM P550Q 1.5GHz POWER5+ server at 55% less price/performance. In addition, the Sun SPARC Enterprise T5220 server has 2.9x better power-performance and has 5.7x better SWaP.

The Sun SPARC Enterprise T5220 performed 2.3x (!) faster than the HP Proliant DL580 and achieved 30% less price/performance. In addition, the Sun SPARC Enterprise T5220 R6iNotes has 4.3x better power-performance and has 8.6x better SWaP.

The Sun SPARC Enterprise T5220 server using a single 1.4GHz UltraSPARC T2 achieved only 22% less performance than 16-core IBM P560Q but Sun has 46% better $/perf. In addition, the Sun SPARC Enterprise T5220 has 2.9x better power-performance and 12x better SWaP.

The SWaP metric is a measure of Server effeciency ratio that includes System performance, power and space consumption on a specific benchmark (SWaP = Performance/[Space(RU)xWatts])

power-performance is computed as Watts/performance. Since power-performance is related to price/performance they are both calculated with performance in the denominator

Lotus Domino 7.0.1 NotesBench R6iNotes Performance Chart (in increasing $/User order)

Users = number of users supported (bigger is better)
NotesMark = the benchmark metric (bigger is better)
$/User = cost per user (smaller is better)

System # Type GHz OS $/User Users Notes
Mark 		#Ptn Resp
Sun Fire
X4200 M2 		2 Opt DC 2.8GHz RedHat Linux4 $2.66 17000 14347 3 496ms
Sun SE T5220 1 US T2 1.4GHz Solaris10 $2.89 43000 36240 6 584ms

HP DL380G5

2 Xeon DC 3GHz Windows 2003 $3.15 19000 15750 3 868 ms

IBM X3650

2 Xeon DC 3GHz SUSE9 Linux $3.47 22000 18989 3 3056ms

Sun Fire
T2000

1 UST1 1.2GHz Solaris10 $3.94 19000 16061 4 400ms

HP Proliant
DL580 G3

4xXeon DC 3GHz Windows 2003 $4.15 18500 15953 4 434ms

Sun Fire
T2000

1 UST1 1.4GHz Solaris10 $4.48 23200 19518 4 692ms

IBM eServer 560Q

4 POWER5+ QC 1.8GHz AIX5L V5.3 $4.89 55000 46193 6 848ms

IBM 550Q

2 POWER5 QC 1.5GHz AIX5L V5.3 $5.97 24000 20108 4 932ms

Sun Fire V890

8 US IV+ DC 1.8GHz Solaris10 $7.19 40000 33862 4 324ms

Complete benchmark results may be found at the Lotus NotesBench website http://www.notesbench.org.

Benchmark Description

The benchmark simulates active users accessing their Domino[R] R6iNotes mail files via standard Web browser. Each simulated user periodically sends, retrieves, and deletes a specified number of e-mail messages from a browser. An average user runs this script four times per hour.

The R5iNotes and R6iNotes workloads, using the Lotus Domino Mail server (R5 or R6) are both HTTP based workloads. R6iNotes is heavier with added features and larger mail files using the MIME format.

The Lotus Webmail and iNotes workloads are NOT comparable.

Disclosure Statement:

NotesBench R6iNotes Sun SPARC Enterprise T5220 (1chip, 8cores/chip@1.4GHz UltraSPARC T2, 8threads/core, 64GB), 6 partitions, Solaris[TM] 10 , Lotus[R] Domino 7.0.1, 43000 users, $2.89per user, 36240 NotesMark tpm, 584ms avg rt. , IBM eServer 550Q, 8x1.5GHz POWER5, 32GB, 4 partitions, AIX 5L V5.3, Lotus Domino 7.0, 24000 users, $5.97 per user, 20108 NotesMark tpm, 932 ms avg rt., HP DL380G5, 2x3.0GHz, Intel Xeon 5160, 6GB, 3 partitions, Windows Server 2003, Lotus[R] Domino 7.0.1, 19000 users, $3.15 per user, 15750 NotesMark tpm, 868ms avg rt., IBM p5 560Q POWER5+ (16x1800 MHz POWER5+), 6 partitions, AIX 5L V5.3, Lotus[R] Domino 7.0.1, 55000 users, $4.89 per user, 46193 NotesMark tpm, 848 ms avg rt. More info: www.notesbench.org

HP DL380G5 power consumption estimated by taking 70% of the maximum input wattage rating of 1193W reported here on 11/14/06. IBM p5 550Q power specifications calculated by applying 70% of the maximum Watts published in "Facts and Features Report", 11/14/06, posted here. Sun SPARC Enterprise T5220 server Power consumption has been taken during the full benchmark execution.

Results Summary

Audited Results
Users: 43000
NotesMark: 36240
Price Performance: $2.89 $/User
Price Performance: $1.186 $ /NotesMark
Response: 584 ms
Systems: One Sun SPARC Enterprise T5220
Number Processors: 1 UltraSPARC T2
Processor GHz: 1.4GHz
Storage: 4xSTK2540, 3xSTK2501 (12x73GB)
Notes Version: Lotus Domino 7.0.1
#Domino Partition 6
Operating System: Solaris 10
Cost: $124,275.44
Other Performance Metrics
Users/CPU: 43000
Users/Core 5375

[2] Comments
Like this post? del.icio.us | furl | slashdot | technorati | digg
Posted at 10:40AM Oct 09, 2007 in Benchmarks
Tags:
Comments:

Hm ... to old results from HP and IBM.
4xXeon DC 3GHz ~= 2x Quard-Core Xeon and is 2 times cheaper T5220 systems and no Power6 results.

Posted by Triffids on October 09, 2007 at 01:08 PM PDT #

These benchmarks are vendor supported, so the more important question is why aren't IBM and X64 vendors posting results on their latest systems?

I hope you are encouraging them to actually post, instead of doing estimation math. This one actually has a regulated $/performance metric.
So if X64 can meet your claims - it is very easy for them to prove it.

Expensive 4-core 4RU IBM p570 power6 is getting beaten by all kinds of UltraSPARC T2 benchmarks and this 4-core IBM p570 is very expensive.

Posted by BM Seer on October 09, 2007 at 02:27 PM PDT #

Post a Comment:
Comments are closed for this entry.