In a previous post, I talked about a meeting I had with a group of Intel sales reps back at the end of June. I’ve gotten a lot of hits on it and I thought I’d post a follow-up. When we met, Intel was in full hype mode about their upcoming solid state drives and promised that they were about to change the market. Fast forward three and a half months and there have been some new developments. While I’m usually a very modest individual, I want to point out that I was right on a few points. Now don’t get me wrong here. I’m not posting out of hubris. I wanted to follow up on this because I think it’s an indicator of greater market conditions. What’s changed is that Intel has lifted the NDA on their first solid state drive.
But I want to work my way back to that. To follow up on my mention of Seagate; at the time (pre stock market crash) Seagate Technology was (and still is) the market leader in magnetic hard drives and I mentioned that its share price looked like a great deal considering the doubt many people had about the realistic timelines surrounding solid state drive technology. Well, then the bottom fell out of the market and everything went to shit. With the economy in its current state it’s hard to see any company actually growing profits and their share prices reflect that. But are Seagate and other hard drive manufacturers facing obsolescence? Not at all. In fact, it’s leadership has just come out with revised roadmaps for 2009 restating that not only will they themselves be releasing their first solid state drives (solely focused on the enterprise market mind you) next year, but that they will continue to lead the expansion of magnetic storage technology in the form of 2TB+ hard drives. Furthermore, they have reiterated that they will be slowly releasing solid state drives into retail as appropriate in light of the fact that they are simply not cost-competitive in the marketplace. By the way, they are speaking directly to Intel with that statement. Not the other SSD manufacturers.
The next thing that happened occurred just yesterday. Apple announced its latest notebook refresh in the way of its new MacBook Pro and MacBook models. And what was the big hardware improvement for these models? Apple completely dropped Intel’s G45 integrated graphics in favor of Nvidia’s latest discrete mobile graphics. Ouch. So basically Apple confirmed to everyone what most already knew, that Intel’s latest graphics chip was not the next big thing in integrated graphics, but rather a barely adequate stop-gap that only served to allow Apple to release its previous generation of notebooks with the stellar Intel mobile CPUs it really wanted.
So we find ourselves back at the point I made almost four months ago. Intel makes simply the best processors on the planet. But they’ve strayed into unfamiliar waters of technology that some of its competitors are far more established in. Consider the following: Intel announced that it would be releasing two models of solid state drives. One mainstream drive based on MLC NAND flash and another enterprise-focused model based on SLC memory. However, only the mainstream MLC drive has actually been tested by independent reviewers, it’s still not available in retail channels, and it’s priced at $600 for a 64GB drive. To put that in perspective, it would be easy to configure a complete Dell laptop with an 80GB hard drive included for just $700. Hmm, $600 on an otherwise inexpensive component that’s not even available yet or $700 on a complete system now?
So did Intel at least get the performance of their drives in line with the hype? Actually, yes. Their mainstream MLC-based drive simply outperforms every other solid state drive currently on the market at any capacity level. Of course Intel knows this and has also priced their drives higher than every other solid state drive currently on the market. Compared to the magnetic hard drives that it is supposedly going to make obsolete, performance is very good but in many applications it’s on par with today’s fastest consumer hard drives and there have been observed performance problems with random write operations. Not exactly the stats you expect to see in a product that proclaims to be the end of magnetic storage technology.
Furthermore, you didn’t see RAID results included in the reviews. Why does that matter? Because many business systems rely on RAID to provide protection from data corruption and drive failure. The price of these drives is already a sticking point. If the RAID performance is no better than current technology at a fraction of the capacity, you can expect business customers to look the other way.
The bottom line, Intel needs to pull back on the hype and modestly release their solid state drives. Solid state drives will have their day. The performance speaks for itself, but whatever level of success they achieve in the market will show you exactly where the technology is positioned and how it will fare against a very scalable magnetic storage technology.
As a footnote: remember the example we all hear about how computers that used to be the size of an entire room are now the size of our iPhones? We are fast-approaching the day when technologists will say, “Enterprise SANs that used to fill half of our data center, now resides in just 4Us of rack space.” Magnetic storage technology will be how we get there.