Archive Home arrow Reviews: arrow Storage arrow Intel 311 Solid State Drive SSDSA2VP020G2E
Intel 311 Solid State Drive SSDSA2VP020G2E E-mail
Reviews - Featured Reviews: Storage
Written by Olin Coles   
Tuesday, 10 May 2011
Table of Contents: Page Index
Intel 311 Solid State Drive SSDSA2VP020G2E
Intel SSD Larson Creek
Features and Specifications
SSD Testing Methodology
AS-SSD Benchmark
ATTO Disk Benchmark
CrystalDiskMark 3.0 Tests
Iometer IOPS Performance
EVEREST Disk Benchmark
HDD vs Hybrid Drive vs SSD
Intel SSD Larson Creek Conclusion

Iometer IOPS Performance

Iometer is an I/O subsystem measurement and characterization tool for single and clustered systems. Iometer does for a computer's I/O subsystem what a dynamometer does for an engine: it measures performance under a controlled load. Iometer was originally developed by the Intel Corporation and formerly known as "Galileo". Intel has discontinued work on Iometer, and has gifted it to the Open Source Development Lab (OSDL). There is currently a new version of Iometer in beta form, which adds several new test dimensions for SSDs.

Iometer is both a workload generator (that is, it performs I/O operations in order to stress the system) and a measurement tool (that is, it examines and records the performance of its I/O operations and their impact on the system). It can be configured to emulate the disk or network I/O load of any program or benchmark, or can be used to generate entirely synthetic I/O loads. It can generate and measure loads on single or multiple (networked) systems.

To measure random I/O response time as well as total I/O's per second, Iometer is set to use 4KB file size chunks over a 100% random sequential distribution at a queue depth of 32 outstanding I/O's per target. The tests are given a 50% read and 50% write distribution. While this pattern may not match traditional 'server' or 'workstation' profiles, it illustrates a single point of reference relative to our product field.

The chart below illustrates combined random read and write IOPS over a 120-second Iometer test phase, where highest I/O total is preferred:

Iometer_Random_4K-IOPS_30QD_Results.png

In our Iometer tests, which use 32 outstanding I/O's per target and a random 50/50 read/write distribution, SandForce SSDs clearly outperform the competition when tested which a larger queue depth. The latest SATA 6Gb/s storage solutions lead the pack, but Intel's Lasron Creek 311 Series SSD follows closely behind thanks to its SLC NAND flash components.

In our next section, we test linear read and write bandwidth performance and compare its speed against several other top storage products using EVEREST Disk Benchmark. Benchmark Reviews feels that linear tests are excellent for rating SSDs, however HDDs are put at a disadvantage with these tests whenever capacity is high.



 

Comments 

 
# willingRealNeil 2011-05-12 10:38
I'm willing to give this a try. I think that Intel is in the process of hitting one out of the park with this new implementation of Hybrid drives.
Report Comment
 
 
# Competitors?Olle P 2011-05-12 23:21
"... our performance rating considers how effective the Intel 311 Series solid state drive performs in operations against direct competitor storage solutions."

This doesn't really hold true, in my opinion. Valid competitors should only be those available at $120 or less, since Larson Creek (shortened "LC" below) is designed for low price. That leaves us with the Lite-On SSD, which doesn't appear in many tests, and the HDD based storages.
____________________ _______

Now, if instead of looking at it as a stand-alone unit I prefer validating it as used for SRT.
- Due to the nature of SRT a minimum requirement to make SRT faster than using the HDD alone is that the SSD used is considerably faster than the HDD in reading data, and *at least* as fast in writing data.
- SRT is also supposed to be a way to improve performance at a low cost, so any SSD used should be very affordable to purchase. Total price is far more relevant than the price/storage ratio.

SSDs based on MLC follow the rule that "more is merrier", meaning that the bigger the storage the faster it becomes. (As I've understood it the theoretical bandwidth is somewhat proportional to the number of storage chips, since they to some extent work in parallel.) Therefore low capacity MLC SSDs are not only cheap, but also very slow. Depending on many factors the write speed might drop to, and possibly below, that of HDDs.
SLC, as used in LC, is considerably more expensive per storage volume, but (as shown in the tests) doesn't have it's speed depend nearly as much on the storage space compared to MLC.
I think that keeping the write speed up was the main reason Intel choose to use SLC instead of MLC for this SSD.

It would be very interesting to see performance comparisons, both as stand-alone and in use with SRT, between LC and cheap MLC drives in the about $100 and 32-64GB region. Will LC outperform a cheaper and bigger but slower(?) MLC drive in SRT use?
Report Comment
 
 
# RE: Competitors?David Ramsey 2011-05-13 07:33
Um, no, Olle, the rule (at least for computer memory systems) is "the SMALLER the storage the faster it becomes." If it was true that bigger==faster, then there would be no need for any kind of caching.

As I noted in the article, it's difficult to reliably and repeatedly benchmark caching systems, especially "smart" ones whose internal algorithms are unknown. SImply running the benchmarks in a different order will affect the results.
Report Comment
 
 
# Bigger = more bandwidth.Olle P 2011-05-13 12:06
I'm pretty sure about this.
Not only do several SSD tests where different storage size drives of same model show this, but also the somewhat related RAM on nVidia's graphics cards. Notice the different RAM bus width between GTX460 in the 768 vs 1024MB versions? An added RAM/NAND chip makes for more access.
Report Comment
 
 
# RE: Bigger = more bandwidth.David Ramsey 2011-05-13 19:20
Well, the graphics card thing is putting memory chips together kinda like RAID 0 for hard drives. The speed of the chips isn't any faster, but you're reading from a group of chips simultaneously.
Report Comment
 
 
# RE: Competitors?Olin Coles 2011-05-13 12:17
It actually depends on architecture. The SandForce SSDs generally get faster with capacity, because they reserve more NAND for the integrated buffer. Others rely on separate buffer DRAM, which can either be the same size across the series or grow with capacity. Essentially, there's no rule for size:speed.

As for Olle's remark about fairly comparing to price-similar competitors, he's dismissed that there are several SSDs that offer a better price per gigabyte than the Intel 311. In fact, the Patriot Torqx 2 (which was included in my results) offers a 32GB version for $90.
Report Comment
 
 
# PricingOlle P 2011-05-16 10:38
"...several SSDs that offer a better price per gigabyte than the Intel 311. In fact, the Patriot Torqx 2 (which was included in my results) offers a 32GB version for $90."

I'm well aware that there are plenty of MLC SSDs that offer better price per GB. Most of them are also considerably more than $100 in price.
The Torqx 2 isn't listed on page 4, so it's unclear to me (and all other readers) if it's the 32GB version that's been tested, or if it's another version with different performance.
Report Comment
 
 
# Typo??Sam 2011-05-13 12:32
Quote from the article on page 2 "The Intel 311 Series SSD is intended to be partnered with a high-capacity hard disk drive, using Intel Smart Response Technology with any Intel 6-Series desktop chipset: Z68, HM67, QM67."

HM67 and QM67 are notebook chipsets. Or did you mean to write H67 and Q67?
Report Comment
 
 
# RE: Typo??Olin Coles 2011-05-13 12:36
The models that Intel specifies in their press document are: Z68, HM67, QM67.
Report Comment
 
 
# actuallyRonob 2011-05-18 16:37
Actually SSD does get faster as it gets bigger .. simply because the smaller SSD are only reading from say four nand chips twice rather than eight nand chips once. Its the number of channels that slows the smaller SSD down. IN fact what you notice is the write speed that suffers much more so than the read speed (it does suffer, but to a much lesser extent). so olle p is right.
Report Comment
 
 
# RE: actuallyOlin Coles 2011-05-18 16:40
You presume that all SSDs use the same architecture, and that all SSDs of larger capacity use more channels than models of a smaller capacity. See the recent SandForce chips for reference.
Report Comment
 

Comments have been disabled by the administrator.

Search Benchmark Reviews
QNAP Network Storage Servers

Follow Benchmark Reviews on FacebookReceive Tweets from Benchmark Reviews on Twitter