The Ultra Ethernet Consortium (UEC) has announced this week that the next-generation interconnection consortium has grown to 55 members. And as the group works towards developing the initial version of their ultra-fast Ethernet standard, they have released some of the first technical details on the upcoming standard.
Formed in the summer of 2023, the UEC aims to develop a new standard for interconnection for AI and HPC datacenter needs, serving as a de-facto (if not de-jure) alternative to InfiniBand, which is largely under the control of NVIDIA these days. The UEC began to accept new members back in November, and just in five months' time it gained 45 new members, which highlights massive interest for the new technology. The consortium now boasts 55 members and 715 industry experts, who are working across eight technical groups.
There is a lot of work at hand for the UEC, as the group has laid out in their latest development blog post, as the consortium works to to build a unified Ethernet-based communication stack for high-performance networking supporting artificial intelligence and high-performance computing clusters. The consortium's technical objectives include developing specifications, APIs, and source code for Ultra Ethernet communications, updating existing protocols, and introducing new mechanisms for telemetry, signaling, security, and congestion management. In particular, Ultra Ethernet introduces the UEC Transport (UET) for higher network utilization and lower tail latency to speed up RDMA (Remote Direct Memory Access) operation over Ethernet. Key features include multi-path packet spraying, flexible ordering, and advanced congestion control, ensuring efficient and reliable data transfer.
These enhancements are designed to address the needs of large AI and HPC clusters — with separate profiles for each type of deployment — though everything is done in a surgical manner to enhance the technology, but reuse as much of the existing Ethernet as possible to maintain cost efficiency and interoperability.
The consortium's founding members include AMD, Arista, Broadcom, Cisco, Eviden (an Atos Business), HPE, Intel, Meta, and Microsoft. After the Ultra Ethernet Consortium (UEC) began to accept new members in October, 2023, numerous industry heavyweights have joined the group, including Baidu, Dell, Huawei, IBM, Nokia, Lenovo, Supermicro, and Tencent.
The consortium currently plans to release the initial 1.0 version of the UEC specification publicly sometime in the third quarter of 2024.
"There was always a recognition that UEC was meeting a need in the industry," said J Metz, Chair of the UEC Steering Committee. "There is a strong desire to have an open, accessible, Ethernet-based network specifically designed to accommodate AI and HPC workload requirements. This level of involvement is encouraging; it helps us achieve the goal of broad interoperability and stability."
While it is evident that then Ultra Ethernet Consortium is gaining support across the industry, it is still unclear where other industry behemoths like AWS and Google stand. While the hardware companies involved can design Ultra Ethernet support into their hardware and systems, the technology ultimately exists to serve large datacenter and HPC system operators. So it will be interesting to see what interest they take in (and how quickly they adopt) the nascent Ethernet backbone technology once hardware incorporating it is ready.
Networking
While the new CAMM and LPCAMM memory modules for laptops have garnered a great deal of attention in recent months, it's not just the mobile side of the PC memory industry that is looking at changes. The desktop memory market is also coming due for some upgrades to further improve DIMM performance, in the form of a new DIMM variety called the Clocked Unbuffered DIMM (CUDIMM). And while this memory isn't in use quite yet, several memory vendors had their initial CUDIMM products on display at this year's Computex trade show, offering a glimpse into the future of desktop memory.
A variation on traditional Unbuffered DIMMs (UDIMMs), Clocked UDIMMs (and Clocked SODIMMs) have been created as another solution to the ongoing signal integrity challenges presented by DDR5 memory. DDR5 allows for rather speedy transfer rates with removable (and easily installed) DIMMs, but further performance increases are running up against the laws of physics when it comes to the electrical challenges of supporting memory on a stick – particularly with so many capacity/performance combinations like we see today. And while those challenges aren't insurmountable, if DDR5 (and eventually, DDR6) are to keep increasing in speed, some changes appear to be needed to produce more electrically robust DIMMs, which is giving rise to the CUDIMM.
Standardized by JEDEC earlier this year as JESD323, CUDIMMs tweak the traditional unbuffered DIMM by adding a clock driver (CKD) to the DIMM itself, with the tiny IC responsible for regenerating the clock signal driving the actual memory chips. By generating a clean clock locally on the DIMM (rather than directly using the clock from the CPU, as is the case today), CUDIMMs are designed to offer improved stability and reliability at high memory speeds, combating the electrical issues that would otherwise cause reliability issues at faster memory speeds. In other words, adding a clock driver is the key to keeping DDR5 operating reliably at high clockspeeds.
All told, JEDEC is proposing that CUDIMMs be used for DDR5-6400 speeds and higher, with the first version of the specification covering speeds up to DDR5-7200. The new DIMMs will also be drop-in compatible with existing platforms (at least on paper), using the same 288-pin connector as today's standard DDR5 UDIMM and allowing for a relatively smooth transition towards higher DDR5 clockspeeds.
Memory
While the new CAMM and LPCAMM memory modules for laptops have garnered a great deal of attention in recent months, it's not just the mobile side of the PC memory industry that is looking at changes. The desktop memory market is also coming due for some upgrades to further improve DIMM performance, in the form of a new DIMM variety called the Clocked Unbuffered DIMM (CUDIMM). And while this memory isn't in use quite yet, several memory vendors had their initial CUDIMM products on display at this year's Computex trade show, offering a glimpse into the future of desktop memory.
A variation on traditional Unbuffered DIMMs (UDIMMs), Clocked UDIMMs (and Clocked SODIMMs) have been created as another solution to the ongoing signal integrity challenges presented by DDR5 memory. DDR5 allows for rather speedy transfer rates with removable (and easily installed) DIMMs, but further performance increases are running up against the laws of physics when it comes to the electrical challenges of supporting memory on a stick – particularly with so many capacity/performance combinations like we see today. And while those challenges aren't insurmountable, if DDR5 (and eventually, DDR6) are to keep increasing in speed, some changes appear to be needed to produce more electrically robust DIMMs, which is giving rise to the CUDIMM.
Standardized by JEDEC earlier this year as JESD323, CUDIMMs tweak the traditional unbuffered DIMM by adding a clock driver (CKD) to the DIMM itself, with the tiny IC responsible for regenerating the clock signal driving the actual memory chips. By generating a clean clock locally on the DIMM (rather than directly using the clock from the CPU, as is the case today), CUDIMMs are designed to offer improved stability and reliability at high memory speeds, combating the electrical issues that would otherwise cause reliability issues at faster memory speeds. In other words, adding a clock driver is the key to keeping DDR5 operating reliably at high clockspeeds.
All told, JEDEC is proposing that CUDIMMs be used for DDR5-6400 speeds and higher, with the first version of the specification covering speeds up to DDR5-7200. The new DIMMs will also be drop-in compatible with existing platforms (at least on paper), using the same 288-pin connector as today's standard DDR5 UDIMM and allowing for a relatively smooth transition towards higher DDR5 clockspeeds.
MemoryAs LPCAMM2 adoption begins, the first retail memory modules are finally starting to hit the retail market, courtesy of Micron. The memory manufacturer has begun selling their LPDDR5X-based LPCAMM2 memory modules under their in-house Crucial brand, making them available on the latter's storefront. Timed to coincide with the release of Lenovo's ThinkPad P1 Gen 7 laptop – the first retail laptop designed to use the memory modules – this marks the de facto start of the eagerly-awaited modular LPDDR5X memory era.
Micron's Low Power Compression Attached Memory Module 2 (LPCAMM2) modules are available in capacities of 32 GB and 64 GB. These are dual-channel modules that feature a 128-bit wide interface, and are based around LPDDR5X memory running at data rates up to 7500 MT/s. This gives a single LPCAMM2 a peak bandwidth of 120 GB/s. Micron is not disclosing the latencies of its LPCAMM2 memory modules, but it says that high data transfer rates of LPDDR5X compensate for the extended timings.
Micron says that LPDDR5X memory offers significantly lower power consumption, with active power per 64-bit bus being 43-58% lower than DDR5 at the same speed, and standby power up to 80% lower. Meanwhile, similar to DDR5 modules, LPCAMM2 modules include a power management IC and voltage regulating circuitry, which provides module manufacturers additional opportunities to reduce power consumption of their products.

Source: Micron LPDDR5X LPCAMM2 Technical Brief
It's worth noting, however, that at least for the first generation of LPCAMM2 modules, system vendors will need to pick between modularity and performance. While soldered-down LPDDR5X memory is available at speeds up to 8533 MT/sec – and with 9600 MT/sec on the horizon – the fastest LPCAMM2 modules planned for this year by both Micron and rival Samsung will be running at 7500 MT/sec. So vendors will have to choose between the flexibility of offering modular LPDDR5X, or the higher bandwidth (and space savings) offered by soldering down their memory.
Micron, for its part, is projecting that 9600 MT/sec LPCAMM2 modules will be available by 2026. Though it's all but certain that faster memory will also be avaialble in the same timeframe.
Micron's Crucial LPDDR5X 32 GB module costs $174.99, whereas a 64 GB module costs $329.99.
MemorySamsung had quietly launched its BM1743 enterprise QLC SSD last month with a hefty 61.44 TB SKU. At FMS 2024, the company had the even larger 122.88 TB version of that SSD on display, alongside a few recorded benchmarking sessions. Compared to the previous generation, the BM1743 comes with a 4.1x improvement in I/O performance, improvement in data retention, and a 45% improvement in power efficiency for sequential writes.
The 128 TB-class QLC SSD boasts of sequential read speeds of 7.5 GBps and write speeds of 3 GBps. Random reads come in at 1.6 M IOPS, while 16 KB random writes clock in at 45K IOPS. Based on the quoted random write access granularity, it appears that Samsung is using a 16 KB indirection unit (IU) to optimize flash management. This is similar to the strategy adopted by Solidigm with IUs larger than 4K in their high-capacity SSDs.
A recorded benchmark session on the company's PM9D3a 8-channel Gen 5 SSD was also on display.
The SSD family is being promoted as a mainstream option for datacenters, and boasts of sequential reads up to 12 GBps and writes up to 6.8 GBps. Random reads clock in at 2 M IOPS, and random writes at 400 K IOPS.
Available in multiple form-factors up to 32 TB (M.2 tops out at 2 TB), the drive's firmware includes optional support for flexible data placement (FDP) to help address the write amplification aspect.
The PM1753 is the current enterprise SSD flagship in Samsung's lineup. With support for 16 NAND channels and capacities up to 32 TB, this U.2 / E3.S SSD has advertised sequential read and write speeds of 14.8 GBps and 11 GBps respectively. Random reads and writes for 4 KB accesses are listed at 3.4 M and 600 K IOPS.
Samsung claims a 1.7x performance improvement and a 1.7x power efficiency improvement over the previous generation (PM1743), making this TLC SSD suitable for AI servers.
The 9th Gen. V-NAND wafer was also available for viewing, though photography was prohibited. Mass production of this flash memory began in April 2024.
StorageDuring the Intel keynote hosted by CEO Pat Gelsinger, he gave the world a glimpse into the Intel Client roadmap until 2026. Meteor Lake launched last year on that roadmap, and Lunar Lake, which we dived into yesterday as Intel disclosed technical details about the upcoming platform. Pat also presented a wafer on stage, Panther Lake, and he gave some additional information about Intel's forthcoming Panther Lake platform, which is expected in 2025.
We covered Intel's initial announcement about the Panther Lake platform last year. It is set to be Intel's first client platform using its Intel 18A node. Aside from once again affirming that things are on track for a 2026 launch, Pat Gelsinger, Intel's CEO, also confirmed that they will be powering on the first 18A wafer for Panther Lake as early as next week.
| Intel CPU Architecture Generations | |||||
| Alder/Raptor Lake | Meteor Lake |
Lunar Lake |
Arrow Lake |
Panther Lake |
|
| P-Core Architecture | Golden Cove/ Raptor Cove |
Redwood Cove | Lion Cove | Lion Cove | Cougar Cove? |
| E-Core Architecture | Gracemont | Crestmont | Skymont | Crestmont? | Darkmont? |
| GPU Architecture | Xe-LP | Xe-LPG | Xe2 | Xe2? | ? |
| NPU Architecture | N/A | NPU 3720 | NPU 4 | ? | ? |
| Active Tiles | 1 (Monolithic) | 4 | 2 | 4? | ? |
| Manufacturing Processes | Intel 7 | Intel 4 + TSMC N6 + TSMC N5 | TSMC N3B + TSMC N6 | Intel 20A + More | Intel 18A + ? |
| Segment | Mobile + Desktop | Mobile | LP Mobile | HP Mobile + Desktop | Mobile? |
| Release Date (OEM) | Q4'2021 | Q4'2023 | Q3'2024 | Q4'2024 | 2025 |
One element to consider from last year is that Lunar Lake is built using TSMC, with the Lunar Lake compute tile with Xe2-LPG graphics on TSMC N3B, and the I/O tile on TSMC N6. Pat confirmed on stage that Panther Lake will be on Intel 18A. Still, he didn't confirm whether the chip will be made purely at Intel, or a mix between Intel and external foundries (ala Meteor Lake). Intel has also yet to confirm the CPU cores to be used, but from what our sources tell us, it sounds like it will be the new Cougar Cove and Darkmont cores.
As we head into the second half of 2024 and after Lunar Lake launches, Intel may divulge more information, including the architectural advancements Panther Lake is expected to bring. Until then, we will have to wait and see.
CPUs
While the new CAMM and LPCAMM memory modules for laptops have garnered a great deal of attention in recent months, it's not just the mobile side of the PC memory industry that is looking at changes. The desktop memory market is also coming due for some upgrades to further improve DIMM performance, in the form of a new DIMM variety called the Clocked Unbuffered DIMM (CUDIMM). And while this memory isn't in use quite yet, several memory vendors had their initial CUDIMM products on display at this year's Computex trade show, offering a glimpse into the future of desktop memory.
A variation on traditional Unbuffered DIMMs (UDIMMs), Clocked UDIMMs (and Clocked SODIMMs) have been created as another solution to the ongoing signal integrity challenges presented by DDR5 memory. DDR5 allows for rather speedy transfer rates with removable (and easily installed) DIMMs, but further performance increases are running up against the laws of physics when it comes to the electrical challenges of supporting memory on a stick – particularly with so many capacity/performance combinations like we see today. And while those challenges aren't insurmountable, if DDR5 (and eventually, DDR6) are to keep increasing in speed, some changes appear to be needed to produce more electrically robust DIMMs, which is giving rise to the CUDIMM.
Standardized by JEDEC earlier this year as JESD323, CUDIMMs tweak the traditional unbuffered DIMM by adding a clock driver (CKD) to the DIMM itself, with the tiny IC responsible for regenerating the clock signal driving the actual memory chips. By generating a clean clock locally on the DIMM (rather than directly using the clock from the CPU, as is the case today), CUDIMMs are designed to offer improved stability and reliability at high memory speeds, combating the electrical issues that would otherwise cause reliability issues at faster memory speeds. In other words, adding a clock driver is the key to keeping DDR5 operating reliably at high clockspeeds.
All told, JEDEC is proposing that CUDIMMs be used for DDR5-6400 speeds and higher, with the first version of the specification covering speeds up to DDR5-7200. The new DIMMs will also be drop-in compatible with existing platforms (at least on paper), using the same 288-pin connector as today's standard DDR5 UDIMM and allowing for a relatively smooth transition towards higher DDR5 clockspeeds.
Memory
0 Comments