Intel Launches Mainstream Optane SSD 800P Series Based On 3D Xpoint Memory (hothardware.com)

MojoKid writes: Intel just launched a new family of consumer-targeted Optane solid state drives today, dubbed the Intel Optane SSD 800P. Unlike Intel Optane Memory sticks, which accelerate hybrid storage configurations with hard drives through intelligent data caching, or Intel's flagship Optane SSD 900P that's aimed squarely at hardcore enthusiasts with big budgets, these M.2 form factor Intel Optane 800P SSDs target the meat of the mobile and desktop markets, with higher capacities than Optane Memory but more affordable pricing than the 900P. In the benchmarks, the Optane SSD 800P series drives offered a mixed-bag of performance, with sequential transfers that top out at about 1.4GB/s, but with small file transfers, 4K random and mixed workloads, latency, and overall QoS looking strong. Intel will initially be offering two drives in the Optane SSD 800P series, with M.2 80mm 58GB and 118GB models. Suggested pricing for the drives is $129 for the 58GB capacity and $199 for the 118GB drive.

Comments

Re: Trash

[guruevi]

Careful there, not sure about these particular ones but Intel's current crop emaciates the Samsung even though on paper and synthetical benchmarks (IOPS and transfer rate) the Samsung does better.

Testing it myself, the Samsung does good until you transfer ~2-3GB and then it drops like a brick to the lower 1000s of IOPS instead of the 100,000 or more it gave you.

The problem there is that Samsung gives you a good RAM cache (backed up with huge capacitors on their DataCenter models) but once you request synced rates or exceed that cache, the controller lags behind. In the mean time, the Intel continues chugging along at 70-90k IOPS.

Re:What is Optane for?

[m.dillon]

Well, NVMe is a multi-queue spec. The best drivers and chipsets for it will assign a command and response queue to each cpu in the system. This allows for both lockless queuing operation as well as polling with no cross-cpu contamination. In this regard, NVMe is far, far superior to AHCI (aka SATA, which has only one queue for multiple targets) and SAS chipsets (which typically are not multi-queue).

At 10uS, though, interrupt overhead (with MSI-X vectoring per-cpu) still yields superior cpu-v-data performance. Interrupt overhead is only around ~1uS or so. Still, its getting close. At lower latencies polling will definitely be a win. But even at 10uS, interrupt driven operation still leaves the cpu with extra clocks to do other work in that it wouldn't have with polling.

Another problem is that NVMe chipsets generally don't have anywhere near the 1023+ queues supported by the spec. They usually come in at no more than 31 queues, which is not enough to assign one to each cpu thread on heftier systems. The chipset spec can support a lot more... in fact, many more MSI-X interrupts can be supported as well, but we just don't see it out in the field.

Most chipsets only offer 8 queues, which is near worthless on modern multi-core cpus.

nvme0: Model SAMSUNG_MZVPV128HDGM-00000 BaseSerial S1XVNYAGA03031 nscount=1
nvme0: Request 64/32 queues, Returns 8/8 queues, rw-sep map (8, 8)
nvme1: Model Samsung_SSD_960_EVO_250GB BaseSerial S3ESNX0J219064Y nscount=1
nvme1: Request 64/32 queues, Returns 8/8 queues, rw-sep map (8, 8)
nvme2: Model INTEL_SSDPEKKW256G7 BaseSerial BTPY64430Q5B256D nscount=1
nvme2: Request 64/32 queues, Returns 8/8 queues, rw-sep map (8, 8)
nvme3: Model TOSHIBA-RD400 BaseSerial Z6TS10AUTPEV nscount=1
nvme3: Request 64/32 queues, Returns 7/7 queues, rw-sep map (7, 7)
nvme4: Model WDC_WDS256G1X0C-00ENX0 BaseSerial 170369420988 nscount=1
nvme4: Request 64/32 queues, Returns 16/16 queues, rw-sep map (16, 16)
nvme5: Model BPX BaseSerial 8B7107720F0823024374 nscount=1
nvme5: Request 64/32 queues, Returns 7/7 queues, rw-sep map (7, 7)
nvme6: Model PLEXTOR_PX-256M8PeG BaseSerial P02652102851 nscount=1
nvme6: Request 64/32 queues, Returns 16/16 queues, rw-sep map (16, 16)

Eventually we'll start to see chipsets that implement closer to the queue limit in the spec, at which point we can theoretically assign a queue pair to every active user thread using the storage. But for now I would be happy if chipsets just gave us enough queues to implement two per cpu thread (for priority separation).

Also, Intel NVMe SSDs are *NORTORIOUSLY* bad in multi-queue configurations. Performance is far poorer than other vendors placed in the same configuration. I think this is rather ironic, actually. Intel markets low latency, but their chipsets can't handle it in the real-life configurations that NVMe was designed for.

-Matt

Re:Trash

[EETech1]

Relative latencies:
SRAM 1X
DRAM 10X
Optane 100X
NAND 100,000X
Rust 10,000,000X

https://hothardware.com/review...

The same gains as going from HDD to SDD (1000X) are realized again going from SDD to Optane.

Probably one of those things, you don't know why you'd even need it, until you have it, then you won't want to live without it.
Or you're not demanding enough to even notice either way.