Hmmm, i'm wondering if jmicron and sandforce can claw their way back after all their bad former chips. Jmicron with their insanely slow controllers and sandforce with panic mode.
Man, I'm right there with you. I still shudder when I see Jmicron controller BIOS's come up because of their nature to, you know, make RAID arrays disappear and eSATA drives not to detect.
The Sandforce panic nonsense was way out of control with the first generation controller at least (e.g. Vertex 2). The second generation couldn't do TRIM properly because of the compression scheme, no matter how many firmware patches were released.
No, this "1W per chip" figure only applies to the high end and/or high capacity chips of each generation. The DRAM used here can be operated at moderate speed in an energy efficient regime, so I'd expect less power savings than that.
It's going to be a slow year. We need intel's new chipsets and mobos which will come with more M2 and PCIe slots. Till then, everything is pretty much bottlenecked by the SATAIII interface. Theoretically, there's room for improvement in random operations and mixed usage, but I don't think anyone will be able to figure these out within a year.
There are a bunch of motherboards coming out with M.2 ports right on the board, usually sitting between or near the PCI sots. We should see a lot more coming out with Skylake as well.
JMF815 is already a massive failure. A PCIe controller that's limited to x2 speeds and doesn't support NVMe, what's the point? It's like JMicron's ambition is to scrape as close to the bottom of the barrel as possible.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
14 Comments
Back to Article
iwod - Friday, June 12, 2015 - link
What sort of performance lost are we expecting from DRAM-less? We should get much lower power usage ( The DRAM would draw at least 1W? when active ).HexiumVII - Friday, June 12, 2015 - link
Hmmm, i'm wondering if jmicron and sandforce can claw their way back after all their bad former chips. Jmicron with their insanely slow controllers and sandforce with panic mode.Samus - Friday, June 12, 2015 - link
Man, I'm right there with you. I still shudder when I see Jmicron controller BIOS's come up because of their nature to, you know, make RAID arrays disappear and eSATA drives not to detect.Oxford Guy - Saturday, June 13, 2015 - link
The Sandforce panic nonsense was way out of control with the first generation controller at least (e.g. Vertex 2). The second generation couldn't do TRIM properly because of the compression scheme, no matter how many firmware patches were released.Pessimism - Monday, June 15, 2015 - link
Indeed. JMicron provokes my gag reflex.nightbringer57 - Friday, June 12, 2015 - link
Well it could be kind of fine-ish if they learned from the JMF602's inexistant cache....MrSpadge - Friday, June 12, 2015 - link
No, this "1W per chip" figure only applies to the high end and/or high capacity chips of each generation. The DRAM used here can be operated at moderate speed in an energy efficient regime, so I'd expect less power savings than that.bug77 - Friday, June 12, 2015 - link
It's going to be a slow year. We need intel's new chipsets and mobos which will come with more M2 and PCIe slots. Till then, everything is pretty much bottlenecked by the SATAIII interface.Theoretically, there's room for improvement in random operations and mixed usage, but I don't think anyone will be able to figure these out within a year.
SleepyFE - Friday, June 12, 2015 - link
M.2 doesn't work too well in desktop PCs. I am hoping U.2 becomes commonplace soon, or at least sata express.SleepyFE - Friday, June 12, 2015 - link
To clarify: M.2 works, but you need an adapter that takes up your PCIe slot.Etsp - Friday, June 12, 2015 - link
There are a bunch of motherboards coming out with M.2 ports right on the board, usually sitting between or near the PCI sots. We should see a lot more coming out with Skylake as well.The_Assimilator - Friday, June 12, 2015 - link
JMF815 is already a massive failure. A PCIe controller that's limited to x2 speeds and doesn't support NVMe, what's the point? It's like JMicron's ambition is to scrape as close to the bottom of the barrel as possible.DigitalFreak - Friday, June 12, 2015 - link
They've always been that way.Oxford Guy - Saturday, June 13, 2015 - link
I'm very certain that a lot of regular people would be very pleased with the performance of a SATA 1 ssd.