That SR review? When did they become worse that Toms Hardware?
Seriously bench marking a storage device through Hyper-V, and not clarifying if the guest OS had direct/sole access to the device? The SQL transactions per second look like RAM/CPU limited and not storage limited... The WD Blue SSD gets 3109 TPS, and the 980 Pro gets 3160 TPS, the Intel Optane (which is a ideal for SQL Server) doesn't even get a mention? All devices are with ~3% of each other?
And those graphs? Backward lines with some looking like a preschooler drew them? (What did I miss on how to interpret them)?
Oh, is the 980 Pro is TLC or MLC... On Samsungs own spec sheet:
Unleash the power of the Samsung PCIe 4.0 NVMe SSD 980 PRO for your next-level computing.
www.samsung.com
It mentions
3-bit MLC. I was under the impression that 3-bit = TLC unless Samsung are trying to redefine what MLC (which IIRC is 2-bits per cell) is.
Anand's review says TLC, SR mentions MLC ? Surely people how solely review storage would have picked that up and mentioned it in their review?