PAM4 DSPs Battle LPO for OFC Mindshare

Last year, module vendors demonstrated the first 1.6T optical modules, and this year DSP vendors looked ahead to second-generation 1.6T module designs. Whereas the first 1.6T modules connect a 16x100G host interface to 8x200G optics (16:8), next-generation designs will work with forthcoming 200G/lane switch ASICs, as shown in the top row of the figure. Broadcom disclosed its Sian2 1.6T 8:8 DSP at a March investor event, and Marvell followed by announcing its similar Nova 2 at OFC. Not wanting to be left out of the 1.6T landscape, MaxLinear pre-announced Rushmore, which similarly targets 8:8 designs. Although the company withheld product details, it disclosed Samsung Foundry as its manufacturing partner for Rushmore, setting it apart from competitors using TSMC.

Source: Broadcom

Progress on linear pluggable optics (LPO) and other less-than-full-DSP variants was evident at 100G/lane, but vendors also set the stage for 200G/lane. Last November, Credo Semiconductor was first to announce a transmit-only 800G PAM4 DSP for half-retimed modules, which are now known as linear-receive optics (LRO). At OFC, Marvell joined the LRO bandwagon with Spica Gen2-T, a transmit-only version of its 5nm 800G DSP. The company claims the new chip enables sub-8W 800G modules, representing about a 40% power reduction from full-DSP modules. Although Macom demonstrated the components required for 200G/lane LPO, there was a growing consensus that LRO may be required at 200G/lane. In the meantime, the Tx-only 100G/lane DSPs can serve as LRO proof-of-concept vehicles.

The high cost and power of DSPs, combined with customer interest in LPO, has created an opening for startups offering alternative approaches. TeraSignal emerged from stealth at OFC with a unique IC that it calls an intelligent re-driver. Aimed at 400G and 800G LPO modules, the chip is a 100G/lane linear re-driver built in a CMOS process. That process enables added intelligence, such as a digital eye monitor that enables link tuning and monitoring. Another startup, NewPhotonics, demonstrated 200G/lane equalization in the optical domain. In addition to a transmit equalizer, the company's PIC integrates modulators and lasers.

Overall, OFC 2024 proved there's no one-size-fits-all design for PAM4 optics. The massive scale of AI back-end networks has raised the importance of power reduction, particularly for short-reach optics. With the DSP-versus-LPO debate quickly shifting to 200G/lane designs, we can expect new innovations to appear in 2025.

LightCounting published a post-OFC update to its PAM4 and Coherent DSPs report. Subscribers can login to access it: www.lightcounting.com/login


Comments

Post a Comment

Popular posts from this blog

AMD Looks to Infinity for AI Interconnects

Image
With the formal launch of the MI300 GPU, AMD revealed new plans for scaling the multi-GPU interconnects vital to AI-training performance. The company's approach relies on a partner ecosystem, which stands in stark contrast with NVIDIA's end-to-end solutions. The plans revolve around AMD's proprietary Infinity Fabric and its underlying XGMI interconnect. Infinity Fabric Adopts Switching As with its prior generation, AMD uses XGMI to connect multiple MI300 GPUs in what it calls a hive. The hive shares a homogeneous memory space formed by the HBM attached to each GPU. In current designs, the GPUs connect directly using XGMI in a mesh or ring topology. Each MI300X GPU has up to seven Infinity Fabric links, each with 16 lanes. The 4th-gen Infinity Fabric supports up to 32Gbps per lane, yielding 128GB/s of bidirectional bandwidth per link. At the MI300 launch, Broadcom announced that its next-generation PCI Express (PCIe) switch chip will add support for XGMI. At last October'
Read more

NVIDIA Networks NVLink

Image
I attended several sessions at last week's Hot Chips, and NVIDIA's NVSwitch talk was a standout. Ashraf Eassa did a great job of covering the talk's contents in an NVIDIA blog, so I will focus on analysis here.  SuperPOD Bids Adieu to InfiniBand From a system-architecture perspective, the biggest change is extending NVLink beyond a single chassis. NVLink Network is a new protocol built on the NVLink4 link layer. It reuses 400G Ethernet cabling to enable passive-copper (DAC), active-copper (AEC), and optical links. The build its DGX H100 SuperPOD, NVIDIA designed a 1U switch system around a pair of NVSwitch chips. The pod ("scalable unit") includes a central rack with 18 NVLink Switch systems, connecting 32 DGX H100 nodes in a two-level fat-tree topology. This pod interconnect yields 460.8Tbps of bisectional bandwidth. NVLink Network replaces InfiniBand (IB) as the first level of interconnect in DGX SuperPODs. The A100 generation pod uses an IB HDR leaf/spine inte
Read more

NVIDIA Reveals DGX GH200 System Architecture

Image
We industry analysts sometimes get out over our skis when trying to project details of new products. Following NVIDIA's DGX GH200 announcement at Computex, we noticed industry press making the same mistake. Rather than correct our previous NVLink Network post, we'll explain here what we've since learned. Our revelation came when NVIDIA published a white paper titled NVIDIA Grace Hopper Superchip Architecture . The figure below from that paper shows the interconnects at the HGX-module level. The GPU (Hopper) side uses NVLink as a coherent interconnect, whereas the CPU (Grace) side uses InfiniBand, in this case connected through a Bluefield-3 DPU. From a networking perspective, the NVLink and InfiniBand domains are independent, that is, there is no bridging between the two protocols. HGX Grace Hopper Superchip System With NVLink Switch (Source: NVIDIA) The new DGX GH200 builds a SuperPOD based on this underlying module-level architecture. You've probably seen the headlin
Read more