Fujitsu Exec Thumbs Nose at Solid-State Disks (SSDs)


With companies like Super Talent and OCZ pushing solid-state disk pricing into affordable territory , there has been a recent rash of excitement of building up over what the near future might bring. Can we finally expect to get over the performance bottleneck imposed by hard disk drives? Not so fast, says Joel Hagberg, VP of business development at Fujitsu.

In a recent interview , the high level exec played down the current state of flash memory. Even as the latest batch of SSDs tout impressive performance specs to the tune of 120 to 143 MB/sec read speeds and 80 to 93MB/sec writes, Hagberg claims it's more hype than substance. Hagberg says SSDs are " really good if you're reading stuff, but it doesn't work very well for large file reads and large file writes, and it doesn't work well for random writes ." Because of this, the VP notes a sizeable rift between notebook customers' expectations and real-world experiences.

Hagberg didn't stop his criticism with read and write performance, and also attacked power consumption, another area generally seen as a plus with SSDs. He points out the screen, processor, and RAM as the main power sucking culprits in a notebook, and says that any power savings that come from swapping a hard drive for an SSD are negligible.

So where does that leave SSDs? According to Fujitsu, in handheld devices such as MP3 players and cell phones, where the technology belongs.  Because high performance servers perform a lot of writing, flash memory and its limited lifespan isn't seen as a viable replacement for hard disk drives. And though Hagberg does concede that solid-state disks will eventually " become a large portion of the storage solution ," he doesn't see this scenario occurring within the next two years.

Fujitsu isn't considered at the forefront of the SSD market, but Hagberg says he's " not saying something to slander a technology because we don't have a device ." He insists his company could have released a hybrid drive two years ago, but opted not to after analyzing in-laboratory benchmarks. It's hard to argue with some of the points Hagberg brings up, but is he underestimating the technology amid recent advances, or offering a cold, hard reality check?

Around the web