Intel brought us out to their Folsom campus last week for some in-depth product briefings. Much of our briefing is still under embargo, but the portion that officially lifts this morning is the SSD DC P4800X:

optane-4.png

optane-9.png

MSRP for the 375GB model is estimated at $1520 ($4/GB), which is rather spendy, but given that the product has shown it can effectively displace RAM in servers, we should be comparing the cost/GB with DRAM and not NAND. It should also be noted this is also nearly half the cost/GB of the X25-M at its launch. Capacities will go all the way up to 1.5TB, and U.2 form factor versions are also on the way.

For those wanting a bit more technical info, the P4800X uses a 7-channel controller, with the 375GB model having 4 dies per channel (28 total). Overprovisioning does not do for Optane what it did for NAND flash, as XPoint can be rewritten at the byte level and does not need to be programmed in (KB) pages and erased in larger (MB) blocks. The only extra space on Optane SSDs is for ECC, firmware, and a small spare area to map out any failed cells.

Those with a keen eye (and calculator) might have noted that the early TBW values only put the P4800X at 30 DWPD for a 3-year period. At the event, Intel confirmed that they anticipate the P4800X to qualify at that same 30 DWPD for a 5-year period by the time volume shipment occurs.

Read on for more about the SSD DC P4800X (and other upcoming products!)

optane-13.png

Note that the P4800X is able to exceed 500,000 mixed random IOPS at a Queue Depth of only 12, while Intel's own P3700 can not even reach 100,000 at that same QD.

optane-15.png

The predictions I made based on the previous P4800X spec sheet leak were mostly confirmed at the event. Latencies and QoS figures are extremely impressive across the board.

optane-16.png

Intel brought up an interesting metric to show differences between Optane and their own flagship NAND SSD. The above test steadily increases what they call 'write pressure', which is simply an increasing random write load. Note what happens to the read latency of the P3700, while the P4800X stubbornly remains close to the axis throughout.

optane-21.png

All of this great latency does not do any good unless there is an easy way to put it in a server, and Intel is doing their best to do just that. Intel Memory Drive is a hypervisor-like driver that can map P4800X's (multiple are possible) as if they were an extension to system DRAM.

optane-22.png

In the above example, comparing a server with 768GB of DRAM to one with 128GB of DRAM combined with a pair of P4800X's, 80% of the transactions per second were possible (with 1/6th of the DRAM). More impressive was that matrix multiplication of the data saw a 1.1x *increase* in performance. This seems impossible, as Optane is still slower than DRAM, but the key here was that in the case of the DRAM-only configuration, half of the database was hanging off of the 'wrong' CPU (dual Xeon systems have half of the memory connected to each CPU, meaning accessing the other half requires additional hops to get to the CPU doing the work, adding latency to each transaction).

optane-24.png

Many high-end enterprise vendors spoke at the event, and to say they are all excited about the possibilities enabled by Optane is an understatement. We will be getting some testing time with the P4800X shortly and will be reporting on our results very soon, but the P4800X is not the only story here!

stoneybeach.jpg

We were briefed on (and even demoed hands on) Optane Memory at the event, but those NDA’s are not up yet, so I’ll have to direct you to previously published information on that for now.

There is one more thing that I’m allowed to mention here: An Optane Client SSD is coming! I can’t go into details yet, but I can give you a quick history lesson:

Need I say more?