Dual CPU Database Server Comparison
by Johan De Gelas on December 2, 2004 12:11 AM EST- Posted in
- IT Computing
Benchmarks IBM DB2 8.1.3: Intel versus AMD
The first question that most people will ask is, of course, how the best AMD Opteron compares to the newest Intel Xeon "Nocona" CPU. Below is a quick table to refresh your memory and to enable you to compare price/performance:
| Intel Xeon CPUs | Core | L2 cache | L3 cache | x86-64 bit | In Test | Price |
| 3.60 GHz w/ 1M cache 800 MHz FSB (90nm) | Nocona = "Prescott server" | 1 MB | No | Yes | Yes | $851 |
| 3.40 GHz w/ 1M cache 800 MHz FSB (90nm) | Nocona = "Prescott server" | 1 MB | No | Yes | No | $690 |
| 3.20D GHz w/ 1M cache 800 MHz FSB (90nm) | Nocona = "Prescott server" | 1 MB | No | Yes | No | $455 |
| 3 GHz w/ 1M cache 800 MHz FSB (90nm) | Nocona = "Prescott server" | 1 MB | No | Yes | No | $316 |
| 3.20C GHz w/ 2M cache 533 MHz FSB (.13) | Galatin = "P4 EE Server" | 0,5 MB | 2 MB | No | Yes | $1,043 |
| 3.20 GHz w/ 1M cache 533 MHz FSB (.13) | Galatin = "P4 EE Server" | 0,5 MB | 1 MB | No | No | $690 |
| 3.06A GHz w/ 1M cache 533 MHz FSB (.13) | Galatin = "P4 EE Server" | 0,5 MB | 1 MB | No | Yes | $455 |
| 3.06 GHz w/ 512k cache 533 MHz FSB (.13) | Prestonia = "Northwood Server" | 0,5 MB | No | No | Yes | $316 |
| AMD Opteron CPU's | Core | L2 cache | L3 cache | x86-64 bit | In Test | Price |
| Model 250 (2.4 GHz) | Sledgehammer | 1 MB | No | Yes | Yes | $851 |
| Model 248 (2.2 GHz) | Sledgehammer | 1 MB | No | Yes | Yes | $690 |
| Model 246 (2.0 GHz) | Sledgehammer | 1 MB | No | Yes | No | $455 |
| Model 244 (1.8 GHz) | Sledgehammer | 1 MB | No | Yes | No | $316 |
We were also very curious about the Xeon Nocona, as the it brings higher clock speeds, a bigger L2-cache, no L3-cache and a pipeline 11 stages longer than the previous Xeon "Prestonia" and Xeon "Gallatin", which maxed out at 3.2 GHz. The first two features mentioned should boost the performance quite well, while the two last are disadvantages.
We should emphasize that, as we tested with SUSE SLES 8 (kernel 2.4.21), the Xeon Nocona was disadvantaged, since we could not test it in 64-bit mode. We assure you that we will update this report with 2.6 kernel. For now, we decided to give you a full report on SLES 8 and kernel 2.4. (All numbers are expressed in queries per second.)
| Concurrency | Xeon 3.6 GHz | Dual Xeon 3.2 L3 (2MB) | Dual Xeon 3.2 | Dual Xeon 3.06 L3 (1MB) | Dual Xeon 3.06 | Opteron 250 DDR400 32 bit | Dual Opteron 250 DDR 400 64 bit | Dual Opteron 248 DDR 400 64 bit |
| 1 | 55 | 46 | 44 | 43 | 42 | 57 | 61 | 57 |
| 2 | 87 | 74 | 61 | 72 | 61 | 105 | 118 | 107 |
| 5 | 128 | 104 | 100 | 98 | 98 | 123 | 137 | 129 |
| 10 | 136 | 112 | 107 | 105 | 102 | 129 | 145 | 132 |
| 20 | 136 | 113 | 106 | 106 | 104 | 131 | 147 | 132 |
| 35 | 138 | 113 | 106 | 104 | 99 | 133 | 150 | 129 |
| 50 | 138 | 110 | 106 | 102 | 100 | 130 | 145 | 128 |
All concurrency tests below 5 are not reliable enough to make any firm conclusion, especially for the Xeon. The margin of error is somewhat higher, but that is not all.
As the Dual Xeon with Hyperthreading spawns 4 logical CPUs, with a concurrency of 2, it is possible that only one physical CPU is doing all the work. Looking at the numbers and the linux tool top, we feel pretty sure that this is exactly what happens most of the time. Compare Row "5" with "2", and "2" with "1" to see what I mean. Note that the results of rows 10 to 50 do not vary a lot; so, we look at these numbers for our conclusions. In the table below, you can see an overview of how the different CPUs compare in percentages.
| 3.6 vs 3.2 | 2 MB L3-cache vs none | 1 MB L3-cache vs none | Xeon 3.2 vs 3.06 | Xeon 3.2 vs 3.06 (both with L3) | Xeon 3.6 vs Opteron 250 | Opteron 64 bit vs 32 bit |
| 20% | 3% | 1% | 7% | 7% | -4% | 6% |
| 17% | 22% | 18% | 3% | 3% | -17% | 12% |
| 24% | 4% | 1% | 5% | 5% | 5% | 12% |
| 21% | 5% | 3% | 6% | 6% | 6% | 13% |
| 21% | 6% | 2% | 6% | 6% | 3% | 12% |
| 22% | 7% | 5% | 8% | 8% | 3% | 12% |
| 26% | 4% | 2% | 8% | 8% | 7% | 12% |
If we had published a similar report back in August, the Opteron would enjoyed a landslide victory. Luckily for Intel, Nocona is very competitive and is about 5% faster than the Opteron 250.
The gigantic - for x86 - L3-cache can not help the Xeon much. We measured only a 2% to 5% performance boost from the 1 MB L2-cache (at 3.06 GHz), and a 4% to 7% performance boost from the 2 MB L3-cache (at 3.2 GHz). The L3-cache seems to boost performance as much as 5% to 6% clock speed increase - nothing to write home about. So a Xeon "Galatin" 3.2 GHz 2 MB L3-cache performs more or less like a Xeon "Galatin" 3.4 GHz, if such a beast should exist.
A comparison between the 3.2 GHz and 3.06 GHz shows that CPU clockscaling - given equal cache sizes - is almost perfect, a testimony to how CPU intensive this benchmark is. Clearly, the generalisation, "databases are all about I/O" is not accurate for a number of database applications. Read-heavy databases seem to be "all about the CPU".
Using a 64 bit database (DB2 8.1.3) on a 64 bit operating system delivers about 12% to 13% better performance. Since we didn't use more than 2 GB, the most likely explanation is the fact that the software can make use of 16 registers instead of 8. We also tested with a twice as large database and 4 GB of RAM, and the results were very similar.
The performance of the Nocona Xeon compared to the older Xeons is also remarkable. The database doesn't mind the longer pipeline and absence of the L3-cache. On the contrary, it performs better than its clock speed indicates, leaving the older 3.2 GHz Xeon (with 2 MB L3 cache!) behind with 21% to 22%, while the Nocona has only a 13% clock speed advantage over the latter. To be honest, we expected Nocona, with its huge branch misprediction penalty, a result of its extremely long pipeline, to scale much worse.
Facebook
Twitter
RSS
46 Comments
View All Comments
Decoder - Friday, December 3, 2004 - link
davesbeer: "MS is a joke".You don't know jack about enterprise IT. Mostly of the financial services industry (FSI) companies run on both UNIX and Windows. Some FSI companies have standardized on .NET and SQL Server. I know this because i work in this industry. MS is no joke. MS.NET is no joke and i can assure you MS SQL Server 2005 is no joke. $ for $, MS products deliver more value and ease of use/development/admin then anything else out there. x86-64 will help MS win over some of the 64-bit enterprise computing deals as well. MS is in the best position ever.
davesbeer - Friday, December 3, 2004 - link
I had great faith in Anandtech... until this article... not the hardware aspect but the software aspect...I never see MySql in competition... MS is a joke... known for cheap but not reliable or scalable.. DB2 is the only competition to Oracle but it is not the same database on differing platforms therefor has huge problems for customers.. Only Oracle allows you to move from one platform to another with minimal changes... Oracle is the leader in the DB market.. Gartner includes NON relational database in IBM's numbers which inflates them. Oracle commands about 70% in the Unix space and quite frankly is retaking significant ground in the Windows space with the low cost SE1 DB options.. Interesting to note that IBM benches its hardware with Oracle and not DB2..... The only thing software correct was the fact that Linux is extremely important to all the vendors and becoming more important to corporations...
Puppetman - Friday, December 3, 2004 - link
#32 - Oracle prohibits you from posting benchmarks in their licensing agreement that you have to sign to get a copy of the software, I believe.I guess this is partly because it's so complicated to set up (MySQL is easier, but tuning is still an issue).
I would have liked to have seen Postgres 7 and 8 tested. PostgresQL has the features of Oracle, and 8.0 has some pretty impressive performance numbers (the optimizer seems to be much better than the 7.4 optimizer, in my limited tests).
David
Puppetman - Friday, December 3, 2004 - link
They used a 32-bit version of MySQL 3.23, when a 64-bit version of 4.0 or 4.1 are available.No statement as to the storage engine used in MySQL (ISAM, MyISAM, InnoDB, BDB, etc), but all the big sites using MySQL (Google, Yahoo, etc) use the InnoDB engine, as it provides ACID transactions, tablespaces, foreign keys, etc.
These tests are like testing a Pentium 4 3.4ghz EE CPU with Windows 98.
mbhame - Friday, December 3, 2004 - link
Would've been nice to see some Oracle and SQL Server benches!lindy - Friday, December 3, 2004 - link
Our people soft environment consists of a application/web server running on Windows 2000 up front, with a SUN UNIX server running Oracle on the back end. In Nov-Dec the database server is busy…review/pay raise time of the year. The rest of the year it hums along.We have mostly two tier applications, WEB or Application up front on a server, database in the back on a server. However lots of turn key solutions like our Crystal Reports server and our Remedy server are an application on top of MS SQL….so essentially database servers.
Exchange is a beast, every user hits our single Exchange 2003 server….1600+ users with a total 300+ gigs of email. You are right it’s basically a database server with the Exchange application sitting on top of it…..there is no way to separate it. Exchange 2003 would be a great test for you as there are lots of load simulators for it out there that can simulate many users pounding it.
Why use a NAS when you have a SAN? Our 2 big Windows storage server 2003 file servers use a SAN for their 2 terabytes of data. These servers are backed up over fiber to a tape silo attached to the SAN. It’s about the fastest backup solution out there today. To your point the data on those file servers are slowly moving to a sharepoint solution which is a WEB server up front, and a big MS SQL database in the back….a pretty big paradigm shift.
Anyhow good article, and happy holidays!!!
tygrus - Friday, December 3, 2004 - link
What was the MySQL scalling like with the Opterons?Other OS's, other DBMS, MS Server 2003, MS SQL ?
Nocona ?
4-way intel vs 4-way AMD ?
While it's nice to isolate the CPU performance, I would like to see some more variety and real life tests for the next edition. Part of a DB server is the IO handling and disk sub-system. Try to set them up with same (best) SCSI drives (SCSI RAID card ? on-board, OEM best or after market?). A few more serach, report, maintenance and data mining tasks would be nice. Capacity and expansion options (and cost) for more disks and backup.
The other thing is that less CPU % usage for a given workload will reduce latency for potentially greater productivity. You don't want a DB server running at >50% for most of the time for speed, reliability, transaction growth, DB growth, emergency capacity. If it was <50% then failure of a CPU or it's associated memory (for Opterons) then the server ccould be run without it. I'm not saying that the system would be limited by disk IO to have that CPU <50% but that the system as a whole would be running at half its peek.
Scali - Friday, December 3, 2004 - link
This is nice, but I still miss a few configurations that I would be interested in... For starters, Xeons running in 64 bit mode... And I also wonder how Windows would perform. Windows may scale quite differently from 1 to 4 or more CPUs, and HyperThreading may have a different impact aswell (especially with Windows XP or 2003, which have special scheduling strategies for HyperThreading).I hope that these will be covered in future benchmarks. They will put these results in a new perspective.
Bonebardier - Friday, December 3, 2004 - link
I know, why don't I give my posts more thought - sorry Anand, I got my Tyan model numbers mixed up! The board used does of course show Opteron off to its best.Here's my sign!
Bonebardier - Friday, December 3, 2004 - link
Yet another AMD vs. Intel review that handicaps the AMD contender unduly - why was the Opteron platform equipped with a K8W, when a K8S Pro would have provided double the memory bandwidth, or have I answered my own question?I'm looking at building an Opteron based server and would never dream of providing it with only a single bank of dual-channel RAM shared between the two, certainly not when a board is available that allows each processor to have it's own bank of DC RAM, which can be shared with the other processor if needed. Database apps are precisely the type of app that would benefit from this.
Come on Anand - give your articles the thought they deserve, unless this one was just an Intel Nocona advert......