Original Link: http://www.anandtech.com/show/4318/intel-roadmap-ivy-bridge-panther-point-ssds
Intel’s Roadmap: Ivy Bridge, Panther Point, and SSDsby Kristian Vättö on May 6, 2011 1:05 AM EST
Last week, we told you about Sandy Bridge-E and X79 chipset. Today, we have a lot of interesting news about other Intel products, including a look at the Ivy Bridge platform and upcoming SSDs. Intel still isn’t taking the wraps off of their Ivy Bridge architecture, but yesterday’s 3D Tri-Gate announcement certainly changes the expectations.
Ivy Bridge will be a die shrink of Sandy Bridge and represents a “tick” in Intel’s product line. That means the micro-architecture is mostly the same as Sandy Bridge, but it will be manufactured using 22nm process instead of 32nm. That will bring lower power consumption and thus less heat while keeping the same performance level. Unfortunately, we don’t have any information on core counts, clock speeds, model numbers, cache sizes, or the IGP. At the same core counts as SNB, we might see some fairly high clock speeds (>4GHz, anyone?) since current quad-core Sandy Bridge CPUs already offer Turbo Boost up to 3.8GHz, and adding additional cache to the design is almost a given. Intel is putting additional resources into their IGP as well, so we expect to see some healthy performance and capability improvements.
|Sandy Bridge versus Ivy Bridge|
|Sandy Bridge||Ivy Bridge|
|Transistor technology||2D (Planar)||3D (Tri-Gate)|
|PCI-Express (version)||16 lanes (2.0)||16 lanes (3.0)|
|Turbo Boost version||2.0||2.0|
|Memory support||Up to DDR3-1333||Up to DDR3-1600|
|Quick Sync||1.0||2.0 (?)|
|IGP shader count||Up to 12 EUs||Up to 16 EUs (?)|
The table above summarizes most of the currently known differences. Ivy Bridge will have enhanced AVX support, the on-die PCI-Express graphics links become version 3.0 instead of SNB’s 2.0, and official memory speed support from the IMC gets bumped to DDR3-1600. While we won’t know about the power consumption until we actually get to test an IB CPU, the roadmap lists TDPs that are the same as SNB (95W, 65W, 45W, 35W). The shrink to 22nm and 3D transistors (FinFET) almost represents a two-node process technology jump, so we expect performance at various power levels to increase quite a bit. A final interesting point for many users is that Ivy Bridge is pin compatible with Sandy Bridge, and it will work on current LGA1155 motherboards with the appropriate chipset and a firmware and BIOS update (H61, H67, P67, and Z68 are capable of support IB). Intel will also launch new 7-series chipsets, which we’ll get into below.
Wrapping up the discussion of improvements, let’s focus on the IGP a bit more. As with Sandy Bridge, we expect Intel will have several IGP variants with Ivy Bridge’s graphics. We don’t know what they will be, but we do know that Intel is calling it their “next Gen Intel HD Graphics” and the core GPU will be DX11 capable. It also looks like Intel will add OpenCL 1.1 support and increase the maximum number of EUs from 12 to 16, though either of those elements may change. Intel lists “Next Gen Quick Sync” as another feature, and with the increase in EU count and additional functionality Ivy Bridge might be double the speed of SNB when it comes to transcoding video.
Panther Point Chipsets
Panther Point is the codename for Intel 7-series chipsets that are set to release simultaneously with Ivy Bridge CPUs. They will come in six flavors, three of which are for the consumer market and three more for the business sector. As noted already, the socket will be the same LGA1155 that Sandy Bridge uses, with pin compatibility on the CPUs and chipsets. The following table summarizes what we know about current and upcoming chipsets; unfortunately, we had to leave Z68 out of the charts as we’re currently under NDA there—and the same goes for Intel SRT; tune in next week for additional information on both.
|Intel Chipset Roadmap|
|Code Name||Panther Point||Panther Point||Panther Point||Patsburg||Panther Point||Panther Point||Panther Point|
|Platform Name||Maho Bay||Maho Bay||Maho Bay||Waimea Bay||Maho Bay||Maho Bay||Maho Bay|
1x16, 2x8, or
|1x16, 2x8 Gen3||1x16 Gen3||2x16 to 4x8 Gen3||1x16, 2x8 Gen3||1x16, 2x8 Gen3||1x16, 2x8 Gen3|
|Intel SRT||Yes||No||Yes||No (?)||Yes||No||No|
|Total USB (USB3) Ports||14 (4)||14 (4)||14 (4)||14 (0)||14 (4)||14 (4)||12 (4)|
|Total SATA (6Gbps)||6 (2)||6 (2)||6 (2)||14 (10)||6 (2)||6 (1)||6 (1)|
|Max Independent Displays||3||3||3||N/A||3||3||3|
The only major differences between the consumer chipsets is in the support for PCIe lane configurations as well as SRT. All are set to be released in 1H’12, but they may not all launch at the same time. For example, with Sandy Bridge Intel released P67 and H67 initially and will follow up with Z68 shortly.
In terms of feature support, Intel will allow overclocking in all non-business Panther Point chipsets and will support the IGP on all the new chipsets. That means there will be no direct replacement for P67 (not that it really matters), and similarly there’s no direct equivalent of the non-overclocking H67. This is good news as quite a few people were annoyed by the lack of a “do everything” chipset for SNB (though Z68 should fix this). The display options are also improved relative to SNB/Cougar Point, and now support up to three independent monitors instead of just two. We suspect most users who want to run that many displays will benefit from a discrete GPU, and it’s also worth noting that with SNB CPUs in 7-series boards you will be restricted to two independent displays.
The biggest improvements are in the I/O segment. One long-awaited feature that’s coming in Panther Point is native USB 3.0 support. Many people were disappointed when Intel decided not to include it in Cougar Point, forcing motherboard manufacturers to use a separate USB 3.0 chip. All Panther Point chipsets will have four USB 3.0 ports which is actually pretty nice, considering that most of the current motherboards only come with two USB 3.0 ports. There are already many USB 3.0 devices (mostly external hard drive bays and flash drives), and with the added bandwidth USB 3.0 offers it’s already moving into the mainstream market. Though some rumors reported Panther Point would include support for Thunderbolt, there is absolutely nothing in the current roadmap to suggest its presence in the 7-series chipsets. There’s always the potential for motherboard makers to use a separate chip to add Thunderbolt, but that could be done with any current platform.
Another interesting I/O update will be PCI-Express. As noted earlier, Ivy Bridge will upgrade PCIe from 2.0 to 3.0, which means twice the bandwidth—or a move from 500MB/s to 1GB/s per lane. This is very good news for users who want maximum graphics performance, meaning SLI or CrossFire, but it also helps with technologies like USB 3.0 and SATA 6Gbps. Sandy Bridge and LGA1156 Nehalems (Lynnfield and Clarkdale) have been somewhat limited by PCIe bandwidth in multi-GPU configurations since they had to share 16 lanes. That meant either one GPU at the full 8GB/s bandwidth, two GPUs at 4GB/s each, or four GPUs with only 2GB/s each (using dual-GPU cards).
We investigated this back when Lynnfield launched; while the performance difference was negligible in most games, there was a noticeable FPS drop in some games, especially in quad-GPU configurations. Of course, today’s GPUs are even more powerful so the limited bandwidth could be a bigger deal if we reran the test with the latest hardware. However, Ivy Bridge should abolish this bottleneck as the bandwidth will be double once more. This means a single x8 PCIe 3.0 slot can provide as much bandwidth as an x16 PCIe 2.0 slot, i.e. 8GB/s. The same applies to quad-GPU configurations as an x4 PCIe 3.0 slot will provide the same 4GB/s bandwidth as x8 PCIe 2.0 slot. This implies that there should no longer be a loss in GPU performance when running multi-GPU configuration on Intel’s mainstream platforms, though obviously you will need a GPU (or other PCIe card) that supports PCIe 3.0 in order to utilize the faster speeds. We don’t have any information yet about AMD’s or NVIDIA’s plans for PCIe 3.0, but historically they have been ready when the motherboard support is there.
The business chipsets support two additional features. Q77, which is the high-end chipset, will support both vPro and SIPP (Stable Image Platform Program). Q75 will not support vPro but it will support SIPP, whereas B75 will not support either of those. All of the business chipsets also include native PCI support, which is still important for many companies that have custom peripherals.
Intel’s SSD Plans
|Intel's SSD Roadmap|
|Currently Shipping||Future Products|
|Code Name||Soda Creek||Postville Refresh||Elmcrest||Larsen Creek||Postville Refresh||Lyndonville||Ramsdale|
|Flash||34nm MLC||25nm MLC||34nm MLC||34nm SLC||25nm MLC||25nm MLC-HET||34nm SLC|
|Interface||SATA 3Gb/s||SATA 3Gb/s||SATA 6Gb/s||SATA 3Gb/s||SATA 3Gb/s||SATA 3Gb/s||PCIe 2.0 (?)|
|Read speed (MB/s)||200||270||500||N/A||270||N/A||N/A|
|Write speed (MB/s)||70||220||315||N/A||220||N/A||N/A|
|4KB read (IOPs)||35000||39500||20000||N/A||39500||N/A||N/A|
|4KB write (IOPs)||6600||23000||8000||N/A||23000||N/A||N/A|
It looks like Intel has learned something from their CPU model name fiascos, though to be fair the SSD lineup naming is quite simple. The 300 series is intended for consumers, the 500 series is for enthusiasts/prosumers, and the 700 series is for enterprise customers. Let’s take each in turn.
The soon-to-be-released 320 series is the same as the 2.5” 320 series, only in a 1.8” form factor. “Lyndonville” and “Ramsdale” are set to replace the X25-E lineup, which is frankly long overdue. Unfortunately, both still list SATA 3Gbps as their interface speed, which further explains why Intel is using someone else’s controller for the 510 series. However, most enterprise customers will be stuck with SATA 3Gbps controllers for a while yet so it’s not as big of a problem. The MLC-HET flash memory is supposedly higher quality MLC for enterprise use. Sadly, we don’t have any further information about MLC-HET and how it compares with regular MLC and SLC.
There aren’t any new 500 series parts, so we’ll move to the 300 series. The 20GB 310 series “Larsen Creek” SSD is a special case intended solely for use with Intel’s SRT. That accounts for its small size as well as the use of SLC flash; we’ll have more information on it in the near future, including a full performance review. Pricing is expected to be relatively low (under $100), so with the appropriate platform it could be ideal for users on a tight budget who still want SSD performance.
As always, Intel has many irons in the fire and most are looking very interesting. From ultra-high-end enthusiast processors to low power Atoms and everything in between, they have something to sell you. They also have companion chipsets, motherboards, SSDs, wireless devices, and other odds and ends to go with their processors. It’s no surprise this business model continues to increase their revenue and net profits every year. We always look forward to the steady march of technology; we don’t yet know what 2013 and beyond will bring, but 2011 and 2012 are looking very strong for Intel. AMD’s Bulldozer and Llano certainly have their work cut out for them, but we’ll see where the chips fall in the next few months.