System IO Benchmarks
HammerDB Database Testing
HammerDB works with many different database suites including Oracle, MS SQL, and MySQL. I opted to use MS SQL, and I had to install Microsoft SQL 2016 to utilize all cores and threads of the CPU. All these tests were done on Server 2012 R2 with all updates. The program allows you to build a database; in this case, I built one with 507 warehouses and created 57 virtual users (one per core, the first virtual user controls the rest).
The point of the benchmark is to hammer the system CPU, memory, and storage. Building the test schema took about 3 hours, the 507 warehouses total to about 30GB of data. You cannot see it in the screenshot since this was taken shortly after starting the benchmark, but memory utilization peaks to around 30GB. It is hard to find the right combination of virtual users and warehouses to produce a constant transaction count, as it is common for the transaction counter to dip and then spike up (not ideal), which it isn't doing.
The system was able to process about 12,586,848 transactions per minute steadily until each virtual user executed 5 million transactions; each transaction is a random SQL query to try and produce real results. Databases are used in everything from your local dentist's office server to Facebook's mega warehouses.
The TYAN S7076 uses a server grade Intel i350 Gbit NIC. I used varying packet sizes from 1K to 64K and ran a looped test with IXIA.
Results were quite good, with performance not dipping until reaching much smaller packet sizes. However, if you are looking to support many clients, I would look to grab a 10Gbit NIC in the form of a mezzanine card for the OCP slot provided. TYAN even offers a list of compatible modules in their manual.
There are no voltage or frequency controls other than manual power state control in the BIOS. Therefore, you cannot overclock the CPU. There are however profiles; "Power", "Balanced Power", "Balanced Performance", and "Performance". You can either have the system balance to save power, or balance to increase performance.
I took power and temperature measurements of the system and CPUs to see what the difference between the "Power" and "Performance" settings provided. There is a difference, and it could be quite large considering servers run 24/7.
PRICING: You can find products similar to this one for sale below.
United States: Find other tech and computer products like this over at Amazon's website.
United Kingdom: Find other tech and computer products like this over at Amazon UK's website.
Canada: Find other tech and computer products like this over at Amazon Canada's website.
Recommended for You
- We at TweakTown openly invite the companies who provide us with review samples / who are mentioned or discussed to express their opinion of our content. If any company representative wishes to respond, we will publish the response here.
Latest News Posts
- Next-gen Hololens to have custom AI chip
- The OnePlus 5 8GB now available for immediate dispatch
- Microsoft Paint killed off with new Creators Update
- The new iPhone 8 might be all about AR
- Apple sued by insurance company for alleged iPhone fire
- Comic-Con 2017 trailer round-up: MCU vs DCEU vs the rest
- graphics card not detected via riser cable
- Maximum ratio (x39) with 4,06GHz for i7 4770 on GA Z87X-UD3H?
- Supermicro SuperO Booster Software Guide
- How to make one network using 2 routers
- BIOSTAR introduces the world's first 8-slot PCI-e mining motherboard with the TB250-BTC+
- HyperX unveils HyperX Alloy Elite and TKL HyperX Alloy FPS Pro mechanical gaming keyboards
- Toshiba Memory Corporation develops world's first 3D flash memory with TSV technology
- ADATA releases XPG GAMMIX line with S10 PCIe Gen3x4 NVMe 1.2 SSD and D10 DDR4