File Copy and Archive Performance

The one area where the i-RAM truly offered impressive performance was when copying files on the i-RAM itself, mainly because a file copy is mostly an I/O bound process.

300MB File Copy
Time in Seconds (Lower is Better)
Gigabyte i-RAM (4GB)
25.25s
Western Digital Raptor (74GB)
77.689s

Copying a 300MB folder containing the Firefox source code from the Raptor to itself took about 77 seconds, yielding just under 4MB/s. Doing the same on the i-RAM took about 25 seconds, resulting in an average transfer rate of about 12MB/s. Note that both the Raptor and the i-RAM were far from their peak theoretical transfer rates, indicating that even the i-RAM is susceptible to some sort of performance overhead.

Next up? Copying a 693MB iso from the drive to itself:

693MB File Copy
Time in Seconds (Lower is Better)
Gigabyte i-RAM (4GB)
6.922s
Western Digital Raptor (74GB)
26.304s

The i-RAM averaged around 100MB/s and copied the file in 6.922 seconds. The Raptor did so in 26.305 seconds at an average of 26.3MB/s.

Finally, we copied our 1.7GB Battlefield 2 install directory:

1.76GB File Copy
Time in Seconds (Lower is Better)
Gigabyte i-RAM (4GB)
31.719s
Western Digital Raptor (74GB)
95.953s

Archive operations are also a lot quicker on the i-RAM. Here's how long it took to create a RAR archive of our Firefox source folder:

WinRAR Archive Creation
Time in Seconds (Lower is Better)
Gigabyte i-RAM (4GB)
57s
Western Digital Raptor (74GB)
70s

Un-archiving a 382MB RAR set provided a much closer competition between the Raptor and the i-RAM:

WinRAR Archive Extraction
Time in Seconds (Lower is Better)
Gigabyte i-RAM (4GB)
15s
Western Digital Raptor (74GB)
19s

i-RAM for Applications Overall Performance
Comments Locked

133 Comments

View All Comments

  • simpletech - Tuesday, July 26, 2005 - link

    I think another possible use (besides certain kinds of servers, like mail servers), is for video capture. The size is a bit small, but if you were capturing segments of footage, it might work. And the price could be reasonable.
  • BikeDude - Tuesday, July 26, 2005 - link

    "but 32-bit Windows can't use more than 4GB of RAM, including the swap file size."

    First of all... "Swap file" is a misnomer. We talked about "swap file" back in the Windows 3.1 days when the OS would swap a process' entire memory space to the *swap* file.

    These days the OS will read/write selected pages of a process' memory from/to the cache manager (who may or may not elect to use the disk to get to the physical pagefile). *Paging*, not "swapping". Executables and libraries are memory mapped and thus start their lives with all pages firmly on disk (so a big executable won't necessarily load slow, but many small DLLs OTOH just might).

    I don't have Windows XP in front of me, but my 32-bit Windows 2003 Standard ed. with 4GB memory and 1GB pagefile certainly doesn't seem affected by the limitation you mention. Enterprise edition can address even more physical memory... Each process is still limited to a 2GB virtual address space though. (32-bit processes marked capable of such will gain a 4GB virtual address space under 64-bit Windows)

    I realise that XPSP2, despite PAE, is limited to 4GB physical memory (http://blogs.msdn.com/carmencr/archive/2004/08/06/...">http://blogs.msdn.com/carmencr/archive/2004/08/06/..., but pagefile as well? Nah, sounds iffy.
  • JarredWalton - Tuesday, July 26, 2005 - link

    Without PAE (or something similar), 32-bit OSes are indeed limited to 4GB of RAM. This is what is being referred to, as PAE is limited to Intel and I don't believe it's available on non-Server versions of Windows. (Correct me if I'm wrong, but PAE is pretty much only on Xeons, right?)

    You're right that it's paging instead of swapping now, but there's really not much difference between the two. Basically, you put data onto the HDD in order to free up physical RAM, on the assumption that the least recently used data that was moved to the HDD won't be accessed again for a while.
  • JarredWalton - Tuesday, July 26, 2005 - link

    Anyway, I've modified the comment to reflect the original intent. If you're running PAE and Server, it's a whole different ball game for high memory systems.
  • Penth - Tuesday, July 26, 2005 - link

    Wow, my friend and I talked about the possibilities for these things several times. But at 3x the initial price and not the performance increase I would have expected, the techie in me is disappointed. My wallet is happy though.
  • StanleyBuchanan - Tuesday, July 26, 2005 - link

    I wonder what the issue is with RAID that Anand comments on.... seems odd that it would behave differently than a HD in this respect and cause problems...


    I would love to have 12gb or more... which is enough for Windows XP, a productivity suite, and a modern game... anything more could be run from NAS
  • Zan Lynx - Sunday, July 31, 2005 - link

    Probably something to do with the PCI bus power. Perhaps two of these cards take more juice than the bus expects to provide while on standby.
  • phaxmohdem - Monday, July 25, 2005 - link

    I saw someone else posting as well, but I would very much like to see some database performance numbers from this device, as well as perhaps a web-serving benchmark.
  • xTYBALTx - Monday, July 25, 2005 - link

    How some FPS benchies?
  • GTMan - Monday, July 25, 2005 - link

    I laughed when I saw that line :) A very interesting device and I look forward to where this goes in the future. Your "Final Words" could use a bit of brevity.

Log in

Don't have an account? Sign up now