Comparison Table, Test Setup, Card Reader
| Manufacturer | Kingston | Lexar | Lexar | PNY |
|---|---|---|---|---|
| Family | Ultimate | Platinum II | Professional UDMA | Optima Pro Ultra High Speed |
| Model Number | CF/16GB-U2 | P-CF8GB-266W-DVDC | ||
| Capacity | 16 GB | 16 GB | 16 GB | 8 GB |
| Rated Performance | 266x | 80x | 300x | 266x |
| Flash Memory Type | SLC | - | - | - |
| Form Factor | CF | CF | CF | CF |
| Manufacturer | Sandisk | Silicon Power | Transcend | Transcend |
|---|---|---|---|---|
| Family | Extreme Ducati Edition | 300X Professional | Extreme Speed | Ultra Speed |
| Model Number | SDCFX4-008G-AD1 | SP032GBCFC300V10 | TS8GCF300 | TS32GCF133 |
| Capacity | 8 GB | 32 GB | 8 GB | 32 GB |
| Rated Performance | 300x | 300x | 300x | 133x |
| Flash Memory Type | - | - | SLC | MLC |
| Form Factor | CF | CF | CF | CF |
Test Setup
| System Hardware | |
|---|---|
| Processor(s) | 2x Intel Xeon Processor (Nocona core) 3.6 GHz, FSB800, 1 MB L2 Cache |
| Platrform | Asus NCL-DS (Socket 604) Intel E7520 Chipset, BIOS 1005 |
| RAM | Corsair CM72DD512AR-400 (DDR2-400 ECC, reg.), 2x 512 MB, CL3-3-3-10 Timings |
| System Hard Drive | Western Digital Caviar WD1200JB 120 GB, 7,200 RPM, 8 MB Cache, UltraATA/100 |
| Mass Storage Controller(s) | Intel 82801EB UltraATA/100 Controller (ICH5) |
| Compact Flash | Addonics ADSACFW SATA |
| Secure Digital HC | Transcend M5 USB |
| Networking | Broadcom BCM5721 On-Board Gigabit Ethernet NIC |
| Graphics Subsystem | On-Board Graphics ATI RageXL, 8 MB |
| System Hardware | |
|---|---|
| Performance Measurements | c't h2benchw 3.6 |
| I/O Performance | IOMeter 2003.05.10 Fileserver-Benchmark, Web server-Benchmark, Workstation-Benchmark |
| System Software and drivers | |
|---|---|
| OS | Microsoft Windows Server 2003 Enterprise Edition, Service Pack 1 |
| Platform Driver | Intel Chipset Installation Utility 7.0.0.1025 |
| Graphics Driver | Default Windows Graphics Driver |
Card Reader: Addonics ADSACFW
Since USB 2.0 card readers aren’t fast enough to support the high throughput of the tested CompactFlash cards, we decided to use a reader device that is based on a SATA interface. USB 2.0 devices typically max out at a bit more than 30 MB/s, while this device supports SATA 1.0a at 150 MB/s. Even though fast hard drives and flash SSDs are faster than this, the bandwidth is enough to test CompactFlash cards, which reached up to 47 MB/s in our tests (for the Kingston Ultimate).
The device, an Addonics ADSACFW, comes installed into a 3.5” drive bay, but Addonics also adds standard and low profile slot covers, which means that the reader can also be installed at the rear of your PC. Of course, it can be hooked up to SATA/300 controllers, it supports hot swapping, and you can boot from CF cards as well. Hot swapping, however, requires controller support.
- Business,
- Sandisk,
- Kingston,
- PNY,
- Lexar,
- Transcend,
- Silicon Power,
- compactflash ,
- card ,
- memory
Latest Memory News
Latest Memory reviews
- 12/27 – Four High-End Quad-Channel DDR3 Memory Kits For X79, Reviewed
- 09/05 – Seven 8 GB DDR3 Memory Kits For Your AMD A75 Motherboard
- 04/18 – Eight 8 GB (2 x 4 GB) Memory Kits For P67 Express, Rounded Up
- 11/23 – Memory Upgrade: Is It Time To Add More RAM?
- 07/17 – How Much Power Does Low-Voltage DDR3 Memory Really Save?
I have two issues with this test that may or may not be relevant. First, isn't it important to compare same size cards when benchmarking continuos read/write? Second isn't just 1GB of ram in the server used to test a limiting factor? And of course more important is there no way to benchmark with the typical write pattern of an SLR?
Ok after reading Robs testing linked by Kjeld Olsen I think tomshardware needs to rethink their recommendation. Card size matters, but more importantly kingston is really slow compared to the fastest cards, somewhere around 50% write performance which shouldn't be a recommended buy. It is bvious that the memory card reader used in this test works very differently from DSLR's.
I only care about two things. Which have SLC and a wear-leveling algorithm. Can the review be updated to include that info for all the CF cards?
In my Nikons (D3, D2xs, D200IR) I only use Lexar 300x or Sandisk Extreme IV. Slower than that and the playback of images and ability to shoot another long burst are affected. The lifetime warranties on the pro Lexar and Sandisk CF's help as well.
The test has a valid point if one is going to use the CF's the same way as a hard drive, outside of a camera. Inside of each different camera the write speeds can vary a lot. Also, for transfering images to a computer, you should test small (jpg) and large (raw) files. I find, as did Rob Galbraith, transfering a load of larger raw files does yield higher transfer rates from my firewire 800 card readers. USB card readers just aren't fast enough when you come back from a job, and want to go to sleep ASAP before the next job.
Ok after reading Robs testing linked by Kjeld Olsen I think tomshardware needs to rethink their recommendation. Card size matters, but more importantly kingston is really slow compared to the fastest cards, somewhere around 50% write performance which shouldn't be a recommended buy. It is bvious that the memory card reader used in this test works very differently from DSLR's.
You need to look a little closer. As Kjeld pointed out, it depends on the camera. On the Sony DSLR the Kingston does quite well. What Robs chart really shows is that the performance will depend greatly on the camera and its internal controller. That said, when it comes time to download your pics to a PC, wouldn't you want something that works fast in the PC? Well, the Kingston does great with that hence the win. Personally I hate Kingston as they've always been over-priced. I'd much rather pick something with acceptable performance that costs half as much.
I sometimes shoot continuous with my D70 and D300, but not enough that this is a limiting factor for me. I do care about how long it takes to download photos to my computer, so the test is perhaps more relevant to me than to someone who does a lot of continuous shooting.
Somethime in April I'm building y own Mini-ITX Nas (waiting for the board I want to become available). I plan to use a SATA CF Reader for my OS boot disk - one for FreeNAS, one for Ubuntu Linux. The CF card sellers (addonics and others) all recommend "industrial" CF cards for these purposes. In this case I'm looking between industrial Transcend and Emphase. It would be nice to have seen some of these "industrial" cards included, to see if they're just a relabel of the exact same thing...
snarfies,
The reason they recommend "industrial" CF cards is for the SLC memory and wear leveling. Same reason that I want to know. So look up those stats when you look at an option. I believe the faster Transcend model fits the bill.
And why aren't companies using the sata adapter to add these to netbooks? It would save at least a hundred dollars, not to mention, would be plenty fast for any netbook, although the throughput is lower, the IO and seek times are far less.
We used a CompactFlash-to-SATA card reader to make sure that there was no bottleneck for the eight contenders from Kingston, Lexar, PNY, San Disk, Silicon Power and Transcend.
I understand not wanting to use a USB adapter, but wouldn't a purely electrical adapter from CF to IDE have been even better? Plug the card direcly into the mother board and see how it handles? I wouldn't think a CF-to-SATA adapter would slow it down too badly, but I would rather see no signal conversion at all.
if where possible to make a RAID 0 to boot windows...
To be really useful, it should have fast 4 Kb random read/write.
During PMA 2009, Pretec unveiled the world's first 666x Compact Flash cards with a read/write speed of 100MB/s. This model is at 64GB. They also have a 100GB model but not as fast.
http://www.dpreview.com/news/0903/ [...] rd666x.asp
This CF stuff on Tom's hardware is a bit outdated if they claim that these cards they've tested are the biggest and fastest cards available.
They also got info incorrect about the SanDisk Extreme IV's. The Ducati versions do promise 45MB/s while the Extreme IV's at the time was rated at 40MB/s but when they pulled the Ducati off of main stream North American market shelves, they also improved the Extreme IV's. All the current Extreme IV's boasts 45MB/s and not 40MB/s.
How do I turn off alerts to this thread? I've never done it before. I don't see an "envelope" icon anywhere.
You need to look a little closer. As Kjeld pointed out, it depends on the camera. On the Sony DSLR the Kingston does quite well. What Robs chart really shows is that the performance will depend greatly on the camera and its internal controller. That said, when it comes time to download your pics to a PC, wouldn't you want something that works fast in the PC? Well, the Kingston does great with that hence the win. Personally I hate Kingston as they've always been over-priced. I'd much rather pick something with acceptable performance that costs half as much.
Because this article specifically says it is aimed at DSLR use and that writing is primary. With the sony kingston does not lag behind that much but still firmly behind, and with the mayor player Nikon and Canon only has half the performance. On a site that recommends overclocking to get 5% more fps I would say it is huge, sandisk has 100% improved writespeed over kingston in the latest Canons and Nikons. or 100% more fps if you will.
One more negative comment: the Addonics reader used in this test has gotten AWFUL review comments for its inconsistent operation with different operating systems (apparently no linux support, problematic Windows support) and, especially, chip sets (improper and/or inconsistent operation with a number of Intel chip sets, inconsistent operation with AMD chip sets, unknown operation with nVidia chip sets). Sadly, I feel that this article is essentially useless.
I have to laugh when I read this test - did Toms Hardware tester ever use a modern DSLR? This test could have been great, but it isn´t... :-(
Recommendaiton for next tests: Table of cards incl. memory type and other specifications. For the read/write/IO tests - maybe redo speed tests for all cards with for example: Canon 5DII, Nikon D700 and a least two card readers (for example a usb2 and a firewire based reader) - now that could be interesting... As it is I feel more confused (from this test..) as to which card would fit my needs...
Anyways - I normally find Thomshardware testing excellent, this test should never have happend...
I was impressed with this review until I read some of the comments that call these results into question. So I checked around. Here are comprehensive results on just about any CF card you would want to buy, and their performance shooting both RAW and JPEG with Canon 5D II.
http://www.robgalbraith.com/bins/c [...] =6007-9784
The Kingston 266x Ultimate is well below the competition, which would be the current San disk Extreme III 30MB/sec. CF Cards and the Lexar Professional 233x. Patrick and Achim I am afraid you are incorrect and recommending the wrong cards to your readers.
A really slacker review and test.
1. Background on why you need a SATA CF reader/converter, and some tests that "show" why the SATA is needed comparing internal and external USB readers to teh Addonics SATA reader.
2. Where are the other fast CF cards, like the Ridata 233x Lightning series--been out about 2 years now--the SanDisk Extreme III cards, at least one SanDisk Ultra II for comparison, and the the Ducati isn't much faster, if any, than the MUCH cheaper SD Extreme III. What about the SanDisk Extreme IV announced in 2006? Are you guys smoking crack?
It seems like you had a couple CF cards lying around your office, decided to roast a cone early on Friday, have some fun, a couple beers, and accidentally published this rubbish.