4 ms·
Can someone explain the benefits of POWER8 as compared to Intel? I though the volume of POWER8 chips being low (as compared to the exceedingly powerful Intel a
by bhouston 13y ago
Can someone explain the benefits of POWER8 as compared to Intel? I though the volume of POWER8 chips being low (as compared to the exceedingly powerful Intel and Arm chips) would mean that innovation in that area would be low as well.
- rwmj 13y agoPOWER is fast. I have remote access to a 64-way POWER7 server through work and it really rocks.
- AnthonyMouse 13y agoThat's the interesting thing about POWER. Uses 250 watts? No problem. Costs $5000? Whatever. They only seem to have one design criteria: It has to be fast.
- StillBored 13y agoThe problem with POWER7 has been the IO connectivity. The use of the GX++ bus has been a huge bottleneck (something they are obviously fixing for POWER8). The theoretical bandwidth on GX++ is 20Gbit and so its basically the equivalent of a single x4 PCIE 2.0 slot. This was borderline bad in 2010 when POWER7 was released, now it looks even worse.
- Sanddancer 13y agoRidiculous parallelism. A POWER8 chip has 12 cores, and each core can handle 8 threads. As a result, these chips can keep the pipeline pretty much always full, and provide massive performance boosts to things like database servers.
- joosters 13y agoSun's Niagara line of processors had similar numbers of threads (64/128), but they often had lacklustre performance. With this amount of parallelism, it becomes very difficult to keep all the threads busy, even for highly scalable programs. You'd get hit by all kinds of problems, like blockages due to memory throughput, or shared resources (IIRC the Niagara threads shared FPUs and other processing blocks, so to get them all running 100% of the time you'd have to manage the workload of your tasks insanely carefully) On top of that, so much stuff just doesn't scale well. On the early Niagaras, even ssh-ing into the machine was noticeably slow. Oh? your crypto doesn't use all 64 threads? Hard luck!
- thrownaway2424 13y agoThe advantage is Google can wave this mobo around in Intel's face to get a better price on xeons.
- protomyth 13y agoOne benefit is that IBM is allowing a direct link to the CPU via their CAPI (Coherence Attach Processor Interface). Currently, Intel has frozen everyone out of using their QPI. This resulted in NVIDIA no longer being able to make chipsets like the Ion. An NVIDIA chipset and GPU would be able to go well beyond what NVIDIA is able to do with Intel chips (limited to PCI hooks).
- zurn 13y agoIBM's semi guys have been doing all kinds of wacky things with low volume chips (and also high-volume, see eg. prev gen consoles). For example their System/360 compatible mainframe CPUs are doing 5.5 GHz now. https://en.wikipedia.org/wiki/IBM_zEC12_(microprocessor) https://en.wikipedia.org/wiki/IBM_zEC12_(microprocessor) Chip design is somewhat like software, adding more people or resources doesn't necessarily make your product superior. Look how long tiny AMD has been putting up a fight.
- KMag 13y agoI used to work on Google's indexing system, and I guess for lot of Google's workloads, total machine performance-per-Watt is the key metric. Many server workloads are largely coordinating the disk controller and the network controller, with maybe some heavy integer and a bit of fp processing in the middle, so your normal pure CPU performance-per-Watt benchmarks can be misleading. ("A supercomputer is a device for turning compute-bound problems into I/O-bound problems." --Seymour Cray, possibly apocryphal) POWER8 uses a lot of Watts, but it also has high I/O throughput, so for many server workloads it could beat Intel chips in full system performance-per-Watt. I would be really interested in seeing an I/O-per-Watt CPU benchmark. Many of Google's workloads are embarrassingly parallel. I used to work on Google's indexing system, and one of the binaries I worked with had a bit over a million threads (across many processes and machines) running at any moment, with most of those threads blocked on I/O. POWER8 has a fair number of hardware threads per core, which should help with the heavily multithreaded style of programming used in many Google projects. Google's datacenters use enormous amounts of power. Several of their locations are former Aluminum smelting plants, because Aluminum smelting also uses enormous amounts of power so the power lines are already in place. My team luckily happened to sit next to one of the guys who designs Google's datacenters and we happened to overhear him say something over the phone about not being able to make sense about enough power to power a small town just disappearing from our usage. One of the guys on my team asked him when this power reduction occurred, and if it had happened before. We worked out that the huge swings in power usage happened when we shut down the prototype for the new indexing system. Most of Google's datacenters are largely empty space because they are limited by the power lines and cooling capacity. In another instance, we discovered a mistake had been made in measuring the maximum power drawn by one of the generations of Google servers. Google had a program designed to max out the systems, and they plugged a server into a power meter wall-wart and ran this program for a while. The maximum power usage was under-estimated due to a combination of the office where the measurement was made being cooler than a datacenter (electrical resistance of most conductors has a positive temperature coefficient in the range of temperatures found in working servers), the machine not being allowed sufficient time to warm up, and/or the indexing system being more highly tuned than the program designed to maximize server utilization. (I like to think it was mostly the latter, but I suspect the first two were the main contributors.) The end result was that cooling was under-provisioned in one of the datacenters. During a heatwave in the area occupied by the datacenter used for most of the indexing process, the datacenter began to overheat, so one of the guys on my team was getting temperature updates every 10 to 15 minutes from a guy actually in the datacenter, and adjusting the number of processes running the indexing system up and down accordingly in order to match indexing speed to the cooling capacity. When you're really truly maxing out that many machines 24/7, some machines will break every day, so the indexing system (as most Google systems) are tolerant of processes just being killed either by software fault, hardware fault, or the cross-machine scheduler. Through a combination of realizing smart phones were really going to take off and they were all ARM powered and realizing how important total machine performance-per-Watt is to some major server purchasers lead me to invest in ARM Holdings in mid 2009. (This does not constitute investment advice. The rest of the market has now realized what I realized in 2009 and I don't feel I have more insight than the market at this point.)